Uploaded by Taner Erdogan

MechanicalPropertiesofStructural Steels

advertisement
NIST
NCSTAR1-3D
Federal Building and Fire Safety Investigation of the
World Trade Center Disaster
Mechanical Properties of Structural
Steels
William E. Luecke
J.
David McColskey
McCowan
Stephen W. Banovic^
Christopher N.
Richard
J.
Fields
Timothy Foecke
Thomas
A. Siewert
Frank W. Gayle
National Institute of Standards and Technology
•
Technology Adminisiralion
•
U
S.
DeparlmenI
of
Comm
NISTNCSTAR1-3D
Federal Building and Fire Safety Investigation of the
World Trade Center Disaster
Mechanical Properties of Structural
Steels
William E. Luecke
J.
David McColskey
Christopher N.
McCowan
Stephen W. Banovic
Richard
J.
Fields*
Timothy Foecke
Thomas
A. Siewert
Frank W. Gayle
Materials Science
and Engineering Laboratory
National Institute of Standards
and Technology
*Retired
September 2005
U.S. Department of
Commerce
Carlos M. Gutierrez, Secretary
Technology Administration
Michelle O'Neill, Acting Under Secretary
for
Technology
National Institute of Standards and Technology
William Jeffrey, Director
Disclaimer No.
1
Certain commercial entities, equipment, products, or materials are identified
in this
document
in
order to describe a
procedure or concept adequately or to trace the history of the procedures and practices used. Such identification is
not intended to imply recommendation, endorsement, or implication that the entities, products, materials, or
equipment are necessarily the best available for the purpose. Nor does such identification imply a finding of fault or
negligence by the National Institute of Standards and Technology.
Disclaimer No. 2
The
policy of NIST is to use the International System of Units (metric units) in all publications.
however, units are presented in metric units or the inch-pound system, whichever is prevalent
In this
in
document,
the discipline.
Disclaimer No. 3
Pursuant to section 7 of the National Construction Safety Team Act, the NIST Director has determined that certain
evidence received by NIST in the course of this Investigation is "voluntarily provided safety-related information" that is
"not directly related to the building failure being investigated" and that "disclosure of that information would inhibit the
voluntary provision of that type of information" (15
USC
7306c).
In addition, a substantial portion of the evidence collected by NIST
provided to NIST under nondisclosure agreements.
in
the course of the Investigation has been
Disclaimer No. 4
NIST takes no position as to whether the design or construction of a WTC building was compliant with any code
due to the destruction of the WTC buildings, NIST could not verify the actual (or as-built) construction, the
properties and condition of the materials used, or changes to the original construction made over the life of the
buildings. In addition, NIST could not verify the interpretations of codes used by applicable authorities in determining
since,
compliance when implementing building codes. Where an Investigation report states whether a system was
designed or installed as required by a code provision, NIST has documentary or anecdotal evidence indicating
whether the requirement was met, or NIST has independently conducted tests or analyses indicating whether the
requirement was met.
Use
No
in
Legal Proceedings
part of
any report
NIST investigation into a structural failure or from an investigation under the
Act may be used in any suit or action for damages arising out of any matter
281a; as amended by P.L. 107-231).
resulting from a
National Construction Safety
Team
mentioned
USC
in
such report (15
National Institute of Standards and Technology National Construction Safety Team Act Report 1-3D
Natl. Inst. Stand. Technol. Natl. Constr. Sfty. Tm. Act Rpt. 1-3D, 322 pages (September 2005)
CODEN: NSPUE2
U.S.
GOVERNMENT PRINTING OFFICE
WASHINGTON: 2005
For sale by the Superintendent of Documents, U.S. Government Printing Office
Fax: (202) 512-2250
Phone: (202) 512-1800
internet: bookstore.gpo.gov
Mail: Stop SSOP, Washington, DC 20402-0001
—
—
Abstract
This report provides five types of mechanical properties for steels from the World Trade Center (WTC):
elastic,
room-temperature
tensile.
Specimens of 29
built
to
tensile,
impact, and elevated-temperature
modulus, E, and Poisson's
Elastic properties include
900 °C. The expression
Behavior for
rate,
different steels representing the 12 identified strength levels in the building as
were characterized.
steels.
room-temperature high strain
for E{T) for
T > 723 °C
is
T < 723 °C
is
ratio, v, for
temperatures up
based on measurements of WTC perimeter column
estimated from literature data.
Room
temperature tensile properties
include yield and tensile strength and total elongation for samples of all grades of steel used in the towers.
The
report provides
model
stress-strain curves for
strain curves, surv iving mill test reports,
recovered
steels, bolts,
each type of steel, estimated from the measured
and historically expected values. With a few exceptions, the
and welds met the specifications they were supplied
measured yield strengths of recovered
steels
were
slightly
strain-rate properties for selected perimeter
and core column
elongation and strain rate sensitivity for rates up to 400
s"'.
impact toughness
at
steels
in testing
steels include yield
all
few
cases, the
methodology. High-
and
tensile strength, total
were evaluated with Charpy
were consistent with other
room temperature of nearly
In a
Measured properties were consistent with
literamre reports on other structural steels. Impact properties
column
to.
lower than specified, probably because of a
combination of mechanical damage, natural variability, and differences
Properties for perimeter and core
stress-
testing.
structural steels
WTC steels tested exceeded
15
ft
lbf at
of the
era.
The
room
temperamre. Elevated-temperature stress-strain curves were collected for selected perimeter and core
column and
truss steels.
The
report presents a
methodology
for estimating high-temperature stress-strain
curves for the steels not characterized based on room-temperature behavior and behavior of other
structural steels
from the
literature.
The measured elevated-temperature
steels is consistent with other structural steels
from
that era.
For the
stress-strain
behavior of
complete constitutive law for creep deformation based on experimental measurements. For the
characterized, the report presents a
Keywords: Creep, high
World Trade
methodology
WTC
truss steels, the report presents a
steels not
for estimating the creep deformation law.
strain rate, high temperature, impact,
modulus, tensile strength, yield strength.
Center.
NISTNCSTAR
1-3D,
WTC
Investigation
ill
Abstract
This page intentionally
IV
left
blank.
NISTNCSTAR
1-3D,
WTC Investigation
1
Table of Contents
Abstract
List
iii
of Figures
ix
List of Tables
xv
List of Acronyms
and Abbreviations
xvii
Preface
xix
Acknowledgments
Executive
Chapter
xxix
Summary
xxxi
1
Introduction
1
.
1
Overview of Report
1.1.1
1
.
1
.2
1.1.3
1
.
1
.4
1.1.5
1
.2
1
1
Elastic Properties (Chapter 2)
1
Room
1
Temperature Tensile Properties (Chapter 3)
High-Strain-Rate Properties (Chapter 4)
3
Impact Properties (Chapter 5)
3
Ele\ ated-Temperature Properties (Chapter 6)
3
Description of the Major Building Components
4
1.2.1
Perimeter Columns
4
1.2.2
Core Columns
6
Flooring System
6
1
.2.3
1
.3
Specimen Nomenclature
7
1
.4
Symbols and Abbreviations
8
Chapter 2
Elastic Properties
11
2.1
Introduction
1
2.2
Experimental Procedure
11
2.3
Elastic Properties (E, v,
G)
for
0<T<723 °C
1
2.4
Elastic Properties (E, v,
G)
for
T>910 °C
13
2.5
Elastic Properties (E, v,
G)
for 723
2.6
Uncertainties
14
2.7
References
14
NIST NCSTAR
1-3D.
WTC
Investigation
°C<T<910 °C
13
v
1
Table of Contents
Chapters
Room-Temperature Tensile Properties
19
3.1
Introduction
19
3.2
Test Procedures
19
3.3
3.4
3.5
3.2.1
Steel
19
3.2.2
Bolts
20
3.2.3
Welds
26
28
Results
3.3.1
Steel
28
3.3.2
Bolts
28
3.3.3
Welds
28
Comparison with Engineering Specifications
33
3.4.1
Steel
33
3.4.2
Bolts
60
3.4.3
Welds
60
Recommended Values
63
3.5.1
Steel
63
3.5.2
Bolts
67
3.5.3
Welds
74
3.6
Summary
75
3.7
References
75
3.7.
1
3.7.2
References Available from Publicly Available Sources
75
References Available from Nonpublic Sources
76
Chapter 4
79
High-Strain-Rate Properties
4.1
Introduction
79
4.2
Test Procedures
79
4.3
vi
4.2.1
High Strain-Rate Tension Tests
79
4.2.2
Analysis of High-Strain-Rate Tension Test Data
8
4.2.3
Kolsky Bar Tests
82
4.2.4
Quasi-Static Compression Tests
84
84
Results
4.3.1
High Strain-Rate Tension Tests
84
4.3.2
High Strain-Rate Kolsky Bar Tests
86
4.3.3
Quasi-Static Compression Tests
88
NISTNCSTAR
1-3D,
WTC Investigation
1
Table of Contents
4.4
Discussion
4.4.
1
4.4.2
90
Calculation of Strain-Rate Sensitivity for Tension Tests
91
Calculation of Strain-Rate Sensitivity for Kolsky Tests
92
4.5
High- Strain-Rate Data Provided
4.6
Comparison with Literature Data
95
4.7
Summary
97
4.8
References
99
94
to the Investigation
Chapter 5
Impact Properties
103
5.1
Introduction
103
5.2
Procedures
103
5.3
Results
105
5.4
5.3.1
Perimeter Columns
105
5.3.2
HAZ Materials from Perimeter Columns
106
5.3.3
Core Columns
106
5.3.4
Trusses
107
5.3.5
Truss Seats
107
5.3.6
Bolts
107
107
Discussion
5.4.1
Perimeter Columns
107
5.4.2
HAZ Materials from Perimeter Columns
109
5.4.3
Core Colunms
109
5.4.4
Trusses
109
5.4.5
Truss Seats
110
5.4.6
Expected Values of Impact Toughness
1
10
5.5
Summary
1 1
5.6
References
1 1
Chapter 6
Elevated Temperature Properties
129
6.1
Introduction
129
6.2
Test procedures
129
6.3
6.2.1
Tensile Tests
129
6.2.2
Creep Tests
130
130
Results
NISTNCSTAR
1-3D.
WTC
Investigation
vii
Table of Contents
6.4
6.3.1
Tensile Tests
130
6.3.2
Creep Tests
130
Recommended
6.4.
1
134
values for steels
A Universal Curve
for Elevated-Temperature Tensile Properties
1
34
6.4.2
Analysis of Tensile Data
136
6.4.3
Estimating Elevated-Temperature Stress-Strain Curves
137
6.4.4
Analysis of Creep Data
149
6.4.5
Recommended Values
155
for Bolts
6.5
Summary
157
6.6
References
158
Chapter?
Summary and
Findings
161
7.1
Summary
161
7.2
Findings
162
Appendix A
Data Tables and Supplemental Figures
Appendix B
Effects of Deformation of Wide-Flange Core
163
Columns on Measured
Yield Strength
Appendix C
Provisional Analysis of High-Rate Data
253
263
Appendix D
Deformation of Steels Used
in
WTC
273
7
Appendix E
Specimen Geometry Effects on High-Rate Tensile Properties
vni
NIST NCSTAR
279
1-3D,
WTC
Investigation
1
List of Figures
Figure P-1
The
.
eight projects in the federal building
and
fire safety investigation
of the
WTC
disaster
xxi
Cross section of a perimeter column; sections with and without spandrels
Figure 1-1.
Figure 1-2.
Characteristic perimeter
Figure 1-3.
column panel
4
components.
illustrating the various
Designations in parentheses refer to the specimen nomenclature of Table 1-1
5
Typical welded box columns and rolled wide-flange shapes used
between the 83rd and 86th floors
6
for core
columns
Figure 1^.
Schematic diagram of a floor truss
Figure 2-1.
Young's modulus
Figure 2-2.
Young's modulus, EiT). and shear modulus, G(T) for 0°C < T< 725°C. Young's
modulus was measured on the WTC steels summarized in Table 2-1. Solid line is
Eq. 2-2. Shear modulus, G, calculated from
and v via Eq. 2-4
7
as a function of temperature
15
Poisson's ratio (v) as a function of temperature. The sohd line
Figure 2-3.
polynomial (Eq. 2-3) for 0 °C <
Figure 2—4.
r<
the
fit
of a 4th order
725 °C
16
Young's modulus data
Fractional error in representing the
is
for the three specimens of
perimeter column steel (Table 2-1) using Eq. 2-2
Figure 3-1.
17
specimen typically used for standard room-temperature quasi-static
Flat tensile
21
tensile tests
Figure 3-2.
Flat tensile
specimen typically used for elevated-temperature
Figure 3-3.
Flat tensile
specimen used for room and elevated-temperamre
specimen used
Figure 3-5.
Flat tensile test
specimen used for some creep and elevated-temperature
Figure 3-6.
Round
tensile
Figure 3-8.
Flat tensile
Figure 3-9.
Heat affected zone tensile
test
the specimen portion that
is in
tensile
Notched
Figure 3-12.
The
NISTNCSTAR
1-3D,
specimen typically used for tensile testing of all-weld metal specimens
The flange
tensile
resistance
WTC
24
specimen used for room-temperature tensile testing
specimen with flange/web weld
is
intact.
The flange
25
the specimen portion that
web machined
is in
flush to the flange
tension
Investigation
25
26
specimens
weld shear strength
24
is
tension
Heat affected zone tensile specimen with weld and
surface.
.
tensile
23
Round
1
23
tests
;
Figure 3-7.
Figure 3-1
22
tests
specimen used for room-temperature and elevated-temperamre
tests
Figure 3-10.
some creep
22
tensile tests
Flat tensile test
for
2
tensile tests
3^.
Figure
16
test (a)
before loading, (b) after failure
27
ix
1
List of Figures
Figure 3-13.
Examples of stress-strain curves for perimeter column, core column, and
In most cases, the strains do not represent failure
Figure 3-14.
Stress-strain curves for notched
Figure 3-15.
Load-displacement curves from four tensile
round bar
tests
truss steels.
29
using the specimens in Fig. 3-11
tests
of A 325
bolts,
30
and the average
curve
Figure 3-16.
3
Schematic diagram of the various definitions of yield behavior
in
mechanical testing
of steel..
35
Figure 3-17.
Methodology
Figure 3-18.
Ratio of measured yield strength to specified
longitudinal tests of perimeter
Figure 3-19.
column
Yield behavior in
tests
column
minimum
yield strength for
all
39
steels
Ratio of measured yield strength to specified
longitudinal tests of core
Figure 3-20.
37
for identifying recovered steels
minimum yield
strength for
all
50
steels
of webs of four wide-flange core columns specified as
F,=36ksi
52
Section of C-88c that
Figure 3-22.
Ratio of measured yield strength to specified
is
56
the source of the specimens
Figure 3-21.
minimum yield
strength for
all
59
longitudinal tests of truss steels
Figure 3-23.
Ratio of measured yield strength to specified
minimum yield
strength for
all
60
longitudinal tests of truss seat steels
Figure 3-24.
Representative tensile stress-strain behavior for perimeter column steels from
flanges, outer webs,
Figure 3-25.
and spandrels (plates
1, 2,
69
and 4)
Representative tensile stress-strain behavior for perimeter column steels from inner
webs
69
(plate 3)
Figure 3-26.
Representative tensile stress-strain behavior for selected core column steels
70
Figure 3-27.
Representative tensile stress-strain behavior for truss steels
70
Figure 3-28.
Representation of deformation that occurs in exposed bolt threads
(left)
and
in bolt
threads coupled with nut threads (right)
71
Figure 3-29.
Tensile strength change of A 325 bolts with exposed thread length
72
Figure 3-30.
Displacement near
Figure 3-31.
Data for load-displacement of A 325 bolts from Fig. 3-15 corrected for
slope from literature
Figure 4-1
.
Figure 4-2.
Figure 4-3.
Specimen used
failure
of A 325 bolts as a ftinction of number of threads exposed
73
initial elastic
74
79
for high-strain-rate tension tests
Schematic diagram of the slack adaptor apparatus
80
Schematic of the procedure for estimating the tensile yield strength when ringing
in
the load signal precludes reliable visual estimation
82
Figure 4-4.
Schematic diagram of Kolsky bar apparatus
82
Figure 4-5.
Oscillograph record of an incident pulse that
and
X
is partially
transmitted to the output bar
partially reflected in the input or incident bar
83
NISTNCSTAR
1-3D,
WTC Investigation
List of
Examples of tensile high-rate
Figure 4-6.
steel
stress-strain curves for F,
M26-C1B1-RP
= 50
column
87
;
Figure 4-7.
Example
Figure 4-8.
Quasi-static compression stress-strain curves for the tests
Figure 4-9.
ksi perimeter
Figures
stress-strain
and
strain rate-strain curves for
Kolsky
87
tests
summarized
in
Table 4-4
Strain rate sensitivity of yield and tensile strength as a function of specified
minimum
yield strength
92
Figure 4-10.
Flow
stress as a function
of strain rate for Kolsky
Figure 4-11.
Flow
stress as a function
of strain rate evaluated
on the
Figure 4-12.
Figure 4-13.
89
93
tests
at different strains for
Kolsky
tests
A 325 bolt
94
Comparison of strain rate
steels from the literature
sensitivities
of NIST
WTC steels to values for structural
96
Total elongation, El„ as a function of strain rate for high-strength, perimeter column
steels
and high-strength
98
steels in the literature
Figure 4-14. Total elongation, El„ as a function of strain rate for low-strength, core column steels
and low-strength
Figure 5-1
An example
.
99
steels in the literature
118
transition curve
Charpy impact specimen geometries and orientations with respect
Figure 5-2.
to the plate rolling
118
direction
Longitudinal and transverse Charpy impact data of samples from the flange and
Figure 5-3.
adjacent
HAZ
of perimeter column
perimeter column
C 0-C
1
1
M
N8-C1M1,
HAZ of
119
Longitudinal and transverse Charpy impact data for the spandrel associated with
Figure 5-4.
perimeter column
N8 and
the
web of wide-flange
core column C-80
Transverse Charpy impact data from samples from perimeter column
N13, and N8, and from floor truss components
Figure 5-5.
Figure 5-6.
Longitudinal Charpy impact data for
Figure 5-7.
Summary
A
120
truss seats
M4,
121
325 bolts
122
dependence of absorbed energy on test temperature for all
perimeter and core column steels. The absorbed energy values of the sub-size
specimens have been corrected using Eq. 5-2 to compare them to data from full-size
(10
Figure 5-8.
the flange and adjacent
1
plot of the
mm by
Summary
10
mm)
plot of the
123
specimens
dependence of absorbed energy on
test
temperature for
all
truss
component and
truss seat steels. The absorbed energy values of the sub-size
specimens have been corrected using Eq. 5-2 to compare them to data from
(10
mm by
10
mm)
frill-size
124
specimens
Figure 5-9. Strength-toughness relationships for several types of structural steels from the
WTC
125
construction era, after Irvine (1969)
Figure 5-10.
Scanning electron micrographs of the fracture surface of a Charpy V-notch
longitudinal specimen orientation from an N8-C1M1 perimeter column (WTC 1-14297-100). (a) ductile dimples (oval feamres) and general surface morphology, (b) low
magnification view of large and small ductile dimples on the fracture surface,
higher magnification view
NISTNCSTAR
1-3D,
WTC
Investigation
(c)
125
xi
List of
Figures
Figure 5-11.
Scanning electron micrographs of the fracture surface of an
column sample showing
and growth
The
Figure 5-12.
at
ductile tearing features that
N8-C1M1
form due
perimeter
to fracture initiation
elongated inclusions and pearlite on planes parallel to the rolling plane.
"ductile dimples" in this case are linear features with a peak-valley
Perspective view of the fracture surface of sample
N8-C1M1 showing
morphology
126
the long peak-
valley features characteristic of the fracture surface for transversely oriented impact
specimens. The green line indicates the topography of the fracture surface
Figure 5-13.
A gray-scale
ClMl
image
(a)
126
and compositional maps from fracture surface of an N8-
perimeter column Charpy V-notch specimen. The relative concentrations of
(b) iron, (c)
manganese, and
(d) sulfur.
The surface of the
"ductile dimple"
littered
is
with the remnants of manganese sulfide inclusions
Figure 5-14.
Figure 6-1.
The
fracture surface of a perimeter truss seat,
127
N13-C3B1
that
was
tested at
room
mode
temperature shows cleavage facets, which indicate a
brittle fracture
Elevated-temperature stress-strain curves. Specimen
N8-C1B1 A-FL
is
127
from a
Fy
60 ksi perimeter column flange plate from WTC 1 column 142 between floors
97-100. Annotations refer to individual test specimen numbers
Figure 6-2.
Creep curves of A 242
represent the
fit
from specimen C-132 at 650 °C. Dashed lines
6-14
using
the parameters in Eqs. 6-16, 6-17, and 6-18.
from Eq.
truss steel
Experimental curves are graphically truncated
Figure 6-3.
= 0.05
132
at s
= 0.05
132
Creep curves of A 242
represent the
fit
truss steel from specimen C-132 at 500 °C. Dashed lines
from Eq. 6-14 using the parameters in Eqs. 6-16, 6-17, and 6-18.
Experimental curves are graphically truncated
Figure 6-5.
at e
Creep curves of A 242 truss steel from specimen C-132 at 600 °C. Dashed lines
represent the fit from Eq. 6-14 using the parameters in Eqs. 6-16, 6-17, and 6-18.
Experimental curves are graphically truncated
Figure 6-4.
131
Creep curves of A 242
at e
= 0.05
133
from specimen C-132 at 400 °C. Solid lines
Dashed lines represent the fit from Eq. 6-14 using
truss steel
represent measured creep strain.
the parameters in Eqs. 6-16, 6-17, and 6-18. Experimental curves are graphically
truncated
Figure 6-6.
at s
= 0.05
Ratio,/ of room- to high-temperature yield strength
The spread of data
individual tests are
solid line
steels,
Figure 6-7.
133
is
(F,.)
for all steels characterized.
room temperature exists because for a given steel, the
normalized to the mean room temperature yield strength. The
at
the expression, Eq. 6-1, developed using literature data
on
structural
which are denoted by the smaller symbols
135
Ratio of room- to high-temperature tensile strength (TS) for the steels in Table 6-1.
The spread of data
individual tests are
solid line
is
room temperature exists because for a given steel, the
normalized to the mean room temperature yield strength. The
at
the expression developed for literature data
on
structural steels, Eq. 6-2,
denoted by the smaller symbols
Figure 6-8.
K(T), Eq. 6-5, for the
steel
Figure 6-9.
n(T), 6-6, for the
with
xii
withF, = 36
A 36
steel
135
of Harmathy (1970), used to model the behavior of
140
ksi
A 36
steel
of Harmathy (1970) used to model the behavior of steel
=36ksi
140
NIST NCSTAR
1
-3D,
WTC Investigation
List of
Figure 6-10.
model (dashed lines) for A 36 steel (F, = 36 ksi nominal) overlaid
on the original data used to generate the model (solid lines). The model (Eqs. 6-4, 65, and 6-6 and Table 6—4) makes essentially identical predictions for
Predictions for the
0°C < T < 300 °C, so only one
line is plotted.
<
for strains in the elastic region (s-
1
.
K(T). Eq. 6-5,for the
A 242 Laclede
Note
model should not be used
shown in this region.
that the
0.003). but the curves are
Instead, elastic lines of the appropriate
Figure 6-1
Figures
modulus should be used
steel
141
used to model the behavior of steel with
F,>36ksi
Figure 6-12.
143
n(T), Eq. 6-6, for the
A 242
Laclede
used to model the behavior of
steel,
steel
with
Fy > 36 ksi
Figtire 6-13.
143
Predictions for the
model (dashed hues)
model
original data used to generate the
be used for
for steel with F,.
(solid lines).
strains in the elastic region {e~~
<
for F,
= 60
ksi perimeter
145
columns
146
the expression for the decrease in yield strength for structural steels in
Example
The correspondence
stress-strain curves for F,
Eq. 6-4
8.
144
Yield point calculated from intersection of appropriate Young's modulus and Eq. 6-4
general Eq. 6-1.
Figure 6-1
in this
is
Example of predicted stress-strain curves
calculated using Eq. 6-4
compared with
Figure 6-17.
shown
modulus should be used
A 242 truss steel.
modeled using the appropriate Young's modulus, while
behavior comes from Eq. 6^
small-strain behavior
the large-strain
Figure 6-16.
on the
model should not
Simulated elevated temperature stress-strain curves for the Laclede
The
Figure 6-15.
ksi overlaid
that the
0.003), but the curves are
region. Instead, elastic lines of the appropriate
Figure 6-14.
> 36
Note
is
= 36
148
within the uncertainty of either expression
ksi
WF core columns calculated using
148
'.
Prediction of the function CfT), Eq. 6-16, from strain rate data for
Figure 6-19.
Variation of the parameter^ with temperature. Solid line
Figure 6-20.
Comparison of high-temperature
yield, F, ,
and
is
A 242 truss steels
Eq. 6-17
tensile, TS, strength for bolt steels
152
153
and
the expression for structural steels in general, Eqs. 6-1 and 6-2. Solid symbols are
bolt steels.
Open symbols
the dashed line
NiSTNCSTAR
1-3D.
WTC Investigation
are 'Tire-resistant" bolt steels. Expression for bolt steels
is
156
xiii
List of
Figures
This page intentionally
xiv
left
blank.
NISTNCSTAR
1-3D,
WTC Investigation
1
List of
Table P-1
.
Federal building and
fire safety investigation
Public meetings and briefings of the
Table 1-1
Specimen nomenclature
for perimeter
Specimen nomenclature
for core
Table 1-2.
of the
WTC disaster
xx
WTC Investigation
Table P-2.
.
Tables
xxiii
column specimens
9
box and wide-flange shapes,
trusses,
and
all
other
specimens
10
Table 1-3.
Mechanical testing definitions used
Table 2-1
Specimen data for Young's modulus (E) determination
15
Results of tensile tests on bolts
3
Table 3-2.
Room-temperature weld properties as measured
32
Table 3-3.
Results of transverse tensile tests on welds from specimen
Table 3-1
.
.
floors 97-100, specified
F, = 60
Fillet
Table 3-5.
Summary of mechanical
N-8
(WTC
1,
column
142,
32
ksi)
weld sizes for various plate thicknesses
Table 3—4.
10
in this report
in the core
box columns
32
properties and chemical compositions for steels from
40
low-strength perimeter columns
Table 3-6.
Summary of mechanical
strength perimeter
Table 3-7.
Summary of mechanical
Yawata
Table 3-8.
properties and chemical compositions for steels
from high43
columns
properties, chemical compositions, and relevant
specifications for steels from high-strength perimeter
Summary of mechanical
ASTM and
45
columns
properties and chemical compositions for steels from core
49
column wide-flange shapes
Table 3-9.
Summary of mechanical
properties and chemical compositions for steels from core
box
columns
51
Table 3-10.
Common truss component dimensions
Table 3-11.
Summary of mechanical
and standards
properties and chemical compositions, and specifications
58
for truss steels tested
Table 3-12.
Summary of mechanical
properties, chemical compositions for truss seat
61
steels tested
Table 3-13.
Estimated
column
Table 3-14.
and
static yield strengths
and work-hardening parameters, Eq. 3-5, for perimeter
66
steels
Estimated
static yield strengths
truss steels
and work-hardening parameters, Eq. 3-5, for core column
68
Table 3-15.
Room-temperature weld metal properties as designed
NIST NCSTAR
1-3D.
WTC
Investigation
57
74
xv
1
List of
Tables
Summary of specimens and
results for high strain-rate tests
of perimeter columns
85
Table 4-2.
Suinmary of specimens and
results for high strain-rate tests
of core columns
86
Table 4-3.
Summary of Kolsky
Table 4-4.
Summary of quasi-static compression
Table 4-5.
Summary of stress-strain
Table 4-6.
Comparison of strain
Table 4-7.
Literature data for strain rate sensitivities of structural steels
Table 5-1
Common
Table 4-1
.
.
88
bar tests
89
tests
rate data plotted in Fig.
rate sensitivity
measured
93
in tension
94
and compression
95
sub-size to full-size upper shelf energy correction factors
Table 5-2.
Summary of Charpy
Table 5-3.
Historical data
Table 6-1
Specimens and locations for high-temperature
.
4-10
115
116
data
on Charpy impact toughness of structural
117
steel
tensile tests with
fiill
stress-strain data.
...
13
Table 6-2.
Values for the parameters in the strength reduction equations (Eqs. 6-1 and 6-2)
Table 6-3.
Property data for the
Table 6-4.
Individual Kj and
Table 6-5.
Values of the parameters of Eqs. 6-5 and 6-6 for
Table 6-6.
Property data for the
Table 6-7.
Individual Ki and
Table 6-8.
Values of the parameters of Eqs. 6-5 and 6-6 for
Table 6-9.
Scaling parameters (Eq. 6-4) for
Table 6-10.
Stress-temperature-strain rate data used to evaluate the parameters of Eq.
Table 6-11.
Sources of bolt data
Table 6-12.
Values for the parameters in the strength reduction equations (Eqs. 6-1 and 6-2) for use
with bolts
157
XVI
n,
A
36
steel reported in
used for steels with F,
A 242 Laclede truss
used for
steels
with
all
Harmathy (1970)
= 36
139
with F,
steel tested as part
F,.
> 36
139
ksi
steels
= 36
ksi
141
of the Investigation
142
142
ksi
steels
136
with F,
> 36
144
ksi
WTC steels
147
6-16
151
155
NISTNCSTAR
1-3D,
WTC Investigation
List of
Acronyms and Abbreviations
Acronyms
AISC
American
AISI
American Iron and Steel
ASTM
ASTM International
AWS
American Welding Society
BPS
Building Performance Study
CVN
Charpy V-notch
FATT
fracture appearance transition temperature
FEMA
Federal Emergency
HAZ
heat-affected zone
HSR
high-rate style
HSLA
high-strength, low-alloy
JIS
Japan Industrial Standard
LERA
Leslie E. Robertson Associates
LRFD
load and resistance factor design
METT
mid-energy transition temperature
NIST
National Institute of Standards and Technology
PANYNJ
Port Authority of New
PC&F
Pacific
PONYA
Port of New
SEAoNY
Structural Engineers Association of
SHCR
Skilling, Helle, Christiansen,
SMA
shielded metal arc
SRS
strain rate sensitivity
use
United States Code
WF
wide-flange (a type of structural steel shape
WTC
World Trade Center
WTC
of Steel Construction
WTC 2
Institute
Management Agency
York and
New Jersey
Car and Foundry
York Authority
World Trade Center
1
NISTNCSTAR
Institute
1
New York
& Robertson
now
usually called a W-shape)
(North Tower)
World Trade Center 2 (South Tower)
1-3D.
WTC
Investigation
XVll
List of Acronyms
and Abbreviations
WTC7
World Trade Center
7
Abbreviations
°C
degrees Celsius
°F
degrees Fahrenheit
|im
micrometer
ft
foot
Fy
yield strength
gal
gallon
GPa
gigapascal; 1x10^
h
hour
in.
inch
L
liter
lb
pound
Ibf
pound force
kip
a force equal to
ksi
1
m
meter
min
minute
mm
millimeter
Mn
magnesium
min
minute
MPa
megapascal; 1x10^
s
second
xvni
1
N/m"
000 pounds
,000 pounds per square inch
NISTNCSTAR
1-3D,
WTC
Investigation
Preface
Genesis of This Investigation
Immediately following the
Federal Emergency
terrorist attack
on the World Trade Center (WTC) on September
Management Agency (FEMA) and
the
planning a building performance study of the disaster. The week of October
search efforts ceased, the Building Performance Study
Team went
to the site
This was to be a brief effort, as the study team consisted of experts
away from
their other professional
May
report in
2002, fulfilling
its
11,
2001, the
American Society of Civil Engineers began
who
7, as
soon as the rescue and
and began
assessment.
largely volunteered their time
commitments. The Building Performance Study
goal "to determine probable failure
its
Team
mechanisms and
damage
future investigation that could lead to practical measures for improving the
issued
its
to identify areas
of
resistance of buildings
against such unforeseen events."
On August
21, 2002, with funding
from the U.S. Congress through
Standards and Technology (NIST) announced
On
disaster.
October
signed into law
.
•
To
Team
The
To
the National Institute of
building and fire safety investigation of the
Team Act
(Public
Law
WTC
107-231), was
WTC Investigation was conducted under the authority of the National
Act.
investigation of the
in\ estigate the
WTC disaster were:
building construction, the materials used, and the technical conditions that
contributed to the outcome of the
•
its
2002, the National Construction Safety
The NIST
Construction Safety
The goals of the
1,
FEMA,
WTC disaster.
serx e as the basis for:
-
Improvements
-
Improved
-
Recommended
-
Improved public
in the
way
buildings are designed, constructed, maintained, and used;
tools and guidance for industry
and safety
officials;
revisions to current codes, standards, and practices; and
safety.
specific objectives were:
1.
Determine
aircraft
2.
why and how WTC and WTC 2 collapsed
why and how WTC 7 collapsed;
1
Determine why the
all
following the
initial
impacts of the
and
injuries
and
fatalities
were so high or low depending on location, including
technical aspects of fire protection, occupant behavior, evacuation, and
emergency
response;
3.
Determine what procedures and practices were used in the design, construction, operation, and
maintenance of WTC
4.
1, 2,
and
7;
and
Identify, as specifically as possible, areas in current building
and
fire
codes, standards, and
practices that warrant revision.
NISTNCSTAR
1-3D,
WTC
Investigation
XIX
.
Preface
NIST
is
purpose of NIST investigations
States,
Technology Administration. The
a nonregulatory agency of the U.S. Department of Commerce's
and the focus
on
is
is
to
improve the safety and
NIST
fact finding.
structural integrity
of buildings
in the
performance and emergency response and evacuation procedures
in the
wake of any building
failure that
has resulted in substantial loss of life or that posed significant potential of substantial loss of life.
does not have the statutory authority to
organizations. Further,
make
findings of fault nor negligence
damages
Law
arising out of any matter
mentioned
in
such report (15
USC
NIST
by individuals or
no part of any report resulting from a NIST investigation
into a building failure or
from an investigation under the National Construction Safety Team Act may be used
for
United
investigative teams are authorized to assess building
in
any
suit or action
281a, as amended by Public
107-231).
Organization of the Investigation
The National Construction Safety Team
Dr.
Arden
Bement,
L.
Jr.,
was
led
for this Investigation, appointed
by Dr.
S.
Shyam
by the then NIST Director,
Sunder. Dr. William L. Grosshandler served as
Associate Lead Investigator, Mr. Stephen A. Cauffman served as Program Manager for Administration,
and Mr. Harold E. Nelson served on the team as a private sector expert. The Investigation included eight
whose
interdependent projects
each of these eight projects
in
is
leaders comprised the remainder of the team.
available at http://wtc.nist.gov.
A detailed description of
The purpose of each project
is
summarized
Table P-1 and the key interdependencies among the projects are illustrated in Fig. P-1
,
Table P-1. Federal building and
fire
safety investigation of the
Technical Area and Project Leader
Analysis of Building and Fire Codes and
Practices; Project Leaders: Dr. H. S.
and Mr. Richard
Lew
W. Bukowski
WTC
disaster.
Project Purpose
Document and analyze
the
code provisions, procedures, and
practices used in the design, construction, operation, and
maintenance of the
structural, passive fire protection,
emergency access and evacuation systems of WTC
Baseline Structural Performance and
Aircraft Impact
Damage
Analysis; Project
Analyze the baseline perfonnance of WTC
and
1
and
1, 2,
and
7.
WTC 2 under
design, service, and abnormal loads, and aircraft impact
damage on
Leader: Dr. Fahim H. Sadek
the structural, fire protection, and egress systems.
Mechanical and Metallurgical Analysis of
Structural Steel; Project Leader: Dr. Frank
W. Gayle
Determine and analyze the mechanical and metallurgical properties
and quality of steel, weldments, and connections from steel
recovered from WTC 1. 2, and 7.
Investigation of Active Fire Protection
Investigate the performance of the active fire protection systems in
Systems; Project Leader: Dr. David
WTC
D. Evans; Dr. William Grosshandler
and
Reconstruction of Thermal and Tenability
Reconstruct the time-evolving temperature, thermal environment,
Enviromnent; Project Leader: Dr. Richard
and smoke movement in WTC 1,2, and 7 for use in evaluating the
structural performance of the buildings and behavior and fate of
occupants and responders.
G.
Gann
Structural Fire
Response and Collapse
Analysis; Project Leaders: Dr. John
L.
Gross and Dr. Therese
P.
McAllister
1,2, and 7 and their role in fire control,
fate
emergency response,
of occupants and responders.
Analyze the response of the WTC towers to fires with and without
aircraft damage, the response of WTC 7 in fires, the performance
of composite steel-trussed floor systems, and determine the most
probable structural collapse sequence for WTC 1. 2, and 7.
Occupant Behavior, Egress, and Emergency
Communications; Project Leader: Mr. Jason
Analyze the behavior and fate of occupants and responders, both
those who survived and those who did not, and the performance of
D. Averill
the evacuation system.
Emergency Response Technologies and
Document
Guidelines; Project Leader: Mr.
of the
terrorist attacks
WTC
7,
Lawson
XX
J.
Randall
emergency responders from the time
1 and WTC 2 until the collapse of
including practices followed and technologies used.
the activities of the
on
WTC
NISTNCSTAR
1-3D,
WTC Investigation
Preface
NIST
WTC
Investigation Projects
Nisr
Figure P-1. The eight projects
in the federal building
investigation of the WTC disaster.
National Construction Safety
The NIST Director
Safety
Team
Act.
initial
fire
safety
Team Advisory Committee
also established an advisory committee as
The
and
mandated under the National Construction
members of the committee were appointed following
a public solicitation.
These were:
•
Paul Fitzgerald. Executive Vice President (retired)
Team Advisory Committee
FM Global, National Construction Safety
Chair
•
John Barsom, President, Barsom Consulting, Ltd.
•
John Bryan, Professor Emeritus, University of Maryland
•
David
•
Glenn Corbett, Professor, John Jay College of Criminal
•
Philip
NISTNCSTAR
Collins, President,
The Preview Group,
Inc.
DiNenno, President, Hughes Associates,
1-3D.
WTC
Investigation
Justice
Inc.
XXI
Preface
•
Robert Hanson, Professor Emeritus, University of Michigan
•
Charles Thornton, Co-Chairman and Managing Principal, The Thomton-Tomasetti Group,
Inc.
•
Kathleen Tiemey, Director, Natural Hazards Research and Applications Information Center,
University of Colorado at Boulder
•
Forman Williams,
Director, Center for
Energy Research, University of California
at
San
Diego
This National Construction Safety
Investigation and
commentary on
Team Advisory Committee
drafts
provided technical advice during the
of the Investigation reports prior
to their public release.
NIST
has benefited from the work of many people in the preparation of these reports, including the National
Team Advisory Committee. The
Construction Safety
content of the reports and recommendations,
however, are solely the responsibility of NIST.
Public Outreach
During the course of this Investigation, NIST held pubhc briefings and meetings
solicit input
(listed in
Table P-2) to
from the public, present preliminary findings, and obtain comments on the direction and
progress of the Investigation from the public and the Advisory Committee.
NIST
maintained a publicly accessible
Web
site
during this Investigation
at http://wtc.nist.gov.
The
site
contained extensive information on the background and progress of the Investigation.
NIST's
The
WTC
Public-Private
collapse of the
Response Plan
WTC buildings has led to broad reexamination of how tall buildings are designed,
constructed, maintained, and used, especially with regard to major events such as
and
fires,
natural disasters,
Reflecting the enhanced interest in effecting necessary change, NIST, with support
terrorist attacks.
from Congress and the Administration, has put
in place a
program, the goal of which
is
to
develop and
implement the standards, technology, and practices needed for cost-effective improvements to the safety
and security of buildings and building occupants, including evacuation, emergency response procedures,
and threat mitigation.
The
strategy to
•
A
meet
this
goal
federal building
is
a three-part
and
NIST-led public-private response program
fire safety investigation to
study the most probable factors that
contributed to post-aircraft impact collapse of the
building, and the associated evacuation and
•
A research
for cost-effective
that
xxn
improvements
WTC towers and the 47-story WTC 7
emergency response experience.
and development (R&D) program
recommendations resulting from the
to (a) facilitate the
implementation of
WTC Investigation, and (b) provide the technical basis
to national building
enhance the safety of buildings,
that includes:
their occupants,
and
fire
codes, standards, and practices
and emergency responders.
NISTNCSTAR
1-3D,
WTC Investigation
Preface
WTC
Table P-2. Public meetings and briefings of the
Date
Location
June 24. 2002
New York
August 21. 2002
Gaithersburg.
December
Washington.
Cit>'.
Principal
NY
9.
2002
DC
WTC Investigation.
Media
briefing announcing the formal start of the Investigation.
Media
briefing on release of the Public Update and
for photographs
April
New York City, NY
2003
8,
Agenda
Public meeting: Public coinments on the Draft Plan for the
pending
MD
Investigation.
Joint public
NIST
request
and videos.
forum with Columbia University on
first-person
interviews.
April 29-30. 2003
Gaithersburg.
MD
NCST Advisor},'
Committee meeting on plan
for
and progress on
WTC Investigation with a public comment session.
Ma\
7.
New York
2003
August 26-27. 2003
City.
Gaithersburg.
NY
MD
Media
briefing on release of May
NCST Advisor)'
investigation with a public
New York
September 17.2003
City.
NY
Media and
2003 Progress Report.
Committee meeting on
comment
status
of the
WTC
session.
public briefing on initiation of first-person data
collection projects.
December 2-3. 2003
Februar\' 12.
Gaithersburg,
New York
2004
MD
Cit>'.
NY
NCST Advisory
Committee meeting on status and initial results
and release of the Public Update with a public comment session.
Public meeting on progress and preliminary findings with public
comments on issues
recommendations.
June
18.
New York
2004
June 22-23, 2004
City.
Gaithersburg,
NY
MD
to
be considered
Media/public briefing on release of June 2004 Progress Report.
NCST Advisory
Committee meeting on
preliminary findings from the
comment
August 24. 2004
system
October 19-20. 2004
Gaithersburg,
MD
22.
2004
Gaithersburg.
MD
fire resistance test
Underwriters Laboratories,
NCST Advisory
set
November
at
Committee meeting on
NCST Advisoiy
City,
NY
status
Committee discussion on
comment
WTC floor
and near complete
Media and public
sequence for the
draft annual report to
session, and a closed session to
discuss pre-draft recommendations for
New York
of
Inc.
of preliminary findings with a public comment session.
Congress, a public
April 5, 2005
the status of and
WTC Investigation with a public
session.
Public viewing of standard
Northbrook. IL
in formulating final
WTC Investigation.
briefing on release of the probable collapse
WTC towers and draft reports for the projects on
codes and practices, evacuation, and emergency response.
New York
June 23. 2005
City,
NY
Media and public
briefing on release of all draft reports for the
WTC towers and draft recommendations for public comment.
September 12-13.
2005
Gaithersburg.
MD
September 13-15.
2005
Gaithersburg,
MD
NCST Advisory
Committee meeting on disposition of public
comments and update to draft reports for the WTC towers.
WTC Technical Conference for stakeholders and technical
community
for dissemination of findings
and opportunity for public
•
make
and recommendations
comments.
technical
A dissemination
and technical assistance program (DTAP) to (a) engage leaders of the
construction and building community in ensuring timely adoption and widespread use of
proposed changes to practices, standards, and codes resulting from the WTC Investigation
and the R&D program, and (b) provide practical guidance and tools to better prepare facility
owners, contractors, architects, engineers, emergency responders, and regulatory authorities
to
The
to
respond to future disasters.
desired outcomes are to
make
buildings, occupants, and first responders safer in future disaster
events.
NISTNCSTAR
1-3D.
WTC
Investigation
xxiii
Preface
National Construction Safety
Team Reports on
the
WTC
Investigation
A final report on the collapse of the WTC towers is being issued as NIST NCSTAR A companion
report on the collapse of WTC 7 is being issued as NIST NCSTAR A. The present report is one of a set
1
.
1
that provides
more
detailed documentation of the Investigation findings and the
As
technical results were achieved.
such,
it
of Investigation publications
means by which
part of the archival record of this Investigation.
is
of the
full set
NIST
(National Institute of Standards and Technology). 2005. Federal Building
these
The
titles
are:
and Fire Safety
World Trade
Investigation of the World Trade Center Disaster: Final Report on the Collapse of the
NIST NCSTAR
Center Towers.
NIST
MD,
September.
(National Institute of Standards and Technology ). 2006. Federal Building
Investigation of the
NIST NCSTAR
Lew, H.
the
Gaithersburg,
1.
S.,
R.
1
and Fire
Safety
World Trade Center Disaster: Final Report on the Collapse of World Trade Center
A. Gaithersburg,
MD.
W. Bukowski, and N.
J.
.
7.
,
Carino. 2005. Federal Building
and Fire
Safety Investigation of
World Trade Center Disaster: Design, Construction, and Maintenance of Structural and Life Safety
Systems.
NIST NCSTAR
1-1.
National Institute of Standards and Technology. Gaithersburg,
MD,
September.
and Fire
Fanella, D. A., A. T. Derecho, and S. K. Ghosh. 2005. Federal Building
Investigation of the
NIST NCSTAR
Safety
World Trade Center Disaster: Design and Construction of Structural Systems.
1-1 A.
National Institute of Standards and Technology. Gaithersburg,
MD,
September.
Ghosh,
S. K.,
and X. Liang. 2005. Federal Building and Fire Safety Investigation of the World
Trade Center Disaster: Comparison of Building Code Structural Requirements. NIST
NCSTAR
1-lB. National Institute of Standards and Technology. Gaithersburg,
Fanella, D. A., A. T. Derecho, and S. K. Ghosh. 2005. Federal Building
Investigation of the
Systems.
MD,
MD,
and Fire
World Trade Center Disaster: Maintenance and Modifications
NIST NCSTAR
September.
Safety
to Structural
1-lC. National Institute of Standards and Technology. Gaithersburg,
September.
and D. A. Johnson. 2005. Federal Building and Fire Safety Investigation of the World
Trade Center Disaster: Fire Protection and Life Safety Provisions Applied to the Design and
Grill, R. A.,
Consti'uction of World Trade Center
Occupancy. NIST
MD,
NCSTAR
1
1, 2,
and
7
and Post-Construction Provisions Applied after
-ID. National Institute of Standards and Technology. Gaithersburg,
September.
Razza,
J.
C, and
R. A. Grill. 2005. Federal Building
and Fire
Safety Investigation of the World
Trade Center Disaster: Comparison of Codes, Standards, and Practices
Design and Construction of World Trade Center
Institute
Grill, R. A.,
and
MD,
7.
NIST
Use
NCSTAR
at the
Time of the
1-lE. National
September.
D. A. Johnson, and D. A. Fanella. 2005. Federal Building and Fire Safety
Investigation of the
XXIV
I, 2,
of Standards and Teclinology. Gaithersburg,
in
World Trade Center Disaster: Comparison of the 1968 and Current (2003)
NIST NCSTAR
1-3D,
New
WTC Investigation
Preface
York City Building Code Provisions.
Technology. Gaithersburg,
Grill. R. A.,
MD,
NIST NCSTAR
1-lF. National Institute of Standards and
September.
and D. A. Johnson. 2005. Federal Building and Fire Safety Investigation of the World
Trade Center Disaster: Amendments
to the
Fire Protection and Life Safety Provisions of the
York City Building Code by Local Laws Adopted While World Trade Center
Use.
NIST NCSTAR
1, 2,
and
7
New
Were
1-lG. National Institute of Standards and Technology. Gaithersburg,
in
MD,
September.
Grill, R. A.,
and D. A. Johnson. 2005. Federal Building and Fire Safety Investigation of the World
to Fire Protection and Life Safety Systems
Trade Center Disaster: Post-Construction Modifications
of World Trade Center
1
and 2. NIST
Technology. Gaithersburg,
Grill, R. A..
MD,
NCSTAR
1-lH. National Institute of Standards and
September.
D. A. Johnson, and D. A. Fanella. 2005. Federal Building and Fire Safety Investigation
of the World Trade Center Disaster: Post-Construction Modifications to Fire Protection, Life
and Structural Systems of World Trade Center 7. NIST NCSTAR l-ll. National Institute of
Safety,
Standards and Technology. Gaithersburg,
MD,
September.
and D. A. Johnson. 2005. Federal Building and Fire Safety Investigation of the World
Grill, R. A.,
Trade Center Disaster: Design,
World Trade Center
MD,
Gaithersburg,
7.
Installation,
NIST NCSTAR
and Operation of Fuel System
1-1 J.
for
Emergency Power
in
National Institute of Standards and Technology.
September.
Sadek, F. 2005. Federal Building and Fire Safety Investigation of the World Trade Center Disaster:
Baseline Structural Performance and Aircraft Impact
Towers.
NIST NCSTAR
1-2.
Damage Analysis of the World Trade
Center
National Institute of Standards and Technology. Gaithersburg,
MD,
September.
Faschan,
W.
J.,
and R. B. Garlock. 2005. Federal Building and Fire
Safety^ Investigation
of the
World Trade Center Disaster: Reference Structural Models and Baseline Performance Analysis of
the World Trade Center Towers. NIST NCSTAR 1-2 A. National Institute of Standards and
Technology. Gaithersburg,
MD,
September.
Kirkpatrick, S. W., R. T. Bocchieri, F. Sadek, R. A. MacNeill, S. Holmes, B. D. Peterson,
R.
W.
Cilke, C. Navarro. 2005. Federal Building
and Fire Safety Investigation of the World Trade
World Trade Center Towers, NIST
Center Disaster: Analysis of Aircraft Impacts into the
NCSTAR
1-2B. National Institute of Standards and Technology. Gaithersburg,
MD,
September.
W. Banovic, T. Foecke, C. N. McCowan, T. A. Siewert, and
J. D. McColskey. 2005. Federal Building and Fire Safety Investigation of the World Trade Center
Disaster: Mechanical and Metallurgical Analysis of Structural Steel. NIST NCSTAR 1-3. National
Gayle, F. W., R.
Institute
J.
Fields.
W.
E.
Luecke,
S.
of Standards and Technology. Gaithersburg,
Luecke,
W.
E., T.
A. Siewert, and F.
Investigation of the
Specifications.
Gaithersburg,
NIST NCSTAR
1-3D,
W.
MD,
September.
Gayle. 2005. Federal Building
and Fire
World Trade Center Disaster: Contemporaneous Structural
NIST
Safety
Steel
Special Publication 1-3 A. National Institute of Standards and Technology.
MD,
September.
WTC
Investigation
XXV
Preface
Banovic,
S.
W.
2005. Federal Building
Disaster: Steel Inventoiy
and Identification.
and Technology. Gaithersburg,
Banovic,
S.
Luecke,
MD,
Safety Investigation of the
NISTNCSTAR
1-3B. National Institute of Standards
September.
Damage and Failure Modes of Structural Steel Components. NIST
1-3C. National Institute of Standards and Technology. Gaithersburg,
W.
E., J.
D. McColskey, C. N.
and
T. A. Siewert,
F.
W.
Gayle. 2005.
McCowan.
W. Banovic,
Federal Building and Fire
Trade Center Disaster: Mechanical Properties of Structural
S.
J.
MD,
September.
Fields, T. Foecke,
Safety Investigation of the
NIST NCSTAR
Steels.
MD,
World
1-3D.
September.
W., C. N. McCowan, and W. E. Luecke. 2005. Federal Building and Fire Safety
Investigation of the
NCSTAR
R.
S.
National Institute of Standards and Technology. Gaithersburg,
Banovic,
World Trade Center
W., and T. Foecke. 2005. Federal Building and Fire Safety Investigation of the World
Trade Center Disaster:
NCSTAR
and Fire
World Trade Center Disaster: Physical Properties of Structural
1-3E. National Institute of Standards and Technology. Gaithersburg,
Evans, D. D., R. D. Peacock, E. D. Kuligowski,
W.
S.
W.
Dols, and
Steels.
MD,
NIST
September.
L. Grosshandler. 2005. Federal
Building and Fire Safety Investigation of the World Trade Center Disaster: Active Fire Protection
Systems.
NIST NCSTAR
1-4.
National Institute of Standards and Technology. Gaithersburg,
MD,
September.
Kuligowski, E. D., D. D. Evans, and R. D. Peacock. 2005. Federal Building and Fire Safety
Investigation of the
World Trade Center Disaster: Post-Construction Fires Prior
NIST NCSTAR
2001.
1-4 A. National Institute of Standards
to
September
and Technology. Gaithersburg,
II,
MD,
September.
Hopkins, M.,
Schoenrock, and E. Budnick. 2005. Federal Building and Fire Safety Investigation
J.
World Trade Center Disaster: Fire Suppression Systems. NIST
of the
Institute
of Standards and Technology. Gaithersburg,
Keough, R.
J.,
and R. A.
Grill.
2005. Federal Building
Trade Center Disaster: Fire Alarm Systems. NIST
and Technology. Gaithersburg,
Ferreira,
M.
J.,
and
S.
M.
MD,
MD,
and Fire
NCSTAR
Safety Investigation of the World
1-4C. National Institute of Standards
September.
Strege. 2005. Federal Building
and Fire
of Standards and Technology. Gaithersburg,
MD,
Safety' Investigation
NCSTAR
Pitts,
in the
the
Ohlemiller,
World Trade Center Towers. NIST
National Institute of Standards and Technology. Gaithersburg,
W.
M., K.
M.
NCSTAR
MD,
NCSTAR
1-5.
September.
and Fire Safety Investigation of
Visual Evidence, Damage Estimates, and Timeline Analysis.
Butler, and V. Junker. 2005. Federal Building
World Trade Center Disaster:
NIST
J.
and K. R. Prasad. 2005. Federal Building and Fire Safety Investigation of the World Trade
Center Disaster: Reconstruction of the Fires
Pitts,
of the
1-4D. National
September.
Gann, R. G., A. Hamins, K. B. McGrattan, G. W. Mulholland, H. E. Nelson, T.
W. M.
1-4B. National
September.
World Trade Center Disaster: Smoke Management Systems. NIST
Institute
NCSTAR
1-5 A. National Institute
of Standards and Technology. Gaithersburg,
MD,
September.
XXVI
NISTNCSTAR
1-3D,
WTC Investigation
Preface
Hamins. A., A. Maranghides. K. B. McGrattan,
J.
Yang. G. MulhoUand, K. R. Prasad,
S.
E. Johnsson, T.
J.
Ohlemiller,
Kukuck, R. Anleitner and
M. Donnelly,
Federal
T. McAllister. 2005.
Building and Fire Safety Investigation of the World Trade Center Disaster: Experiments and
Modeling of Stinctural Steel Elements Exposed to
Standards and Technology. Gaithersburg,
Ohlemiller, T.
J.,
MD,
Fire.
NCSTAR
1-5B. National Institute of
September.
W. Mulholland, A. Maranghides,
G.
NIST
J. J.
Filliben,
and R. G. Gann. 2005. Federal
Building and Fire Safety- Investigation of the World Trade Center Disaster: Fire Tests of Single
NIST NCSTAR
Office Workstations.
Gaithersburg,
MD,
1-5C. National Institute of Standards and Technology.
September.
Gann, R. G., M. A. Riley,
J.
M. Repp, A.
S.
Whittaker, A.
M. Reinhom, and
P.
A. Hough. 2005.
Federal Building and Fire Safety Investigation of the World Trade Center Disaster: Reaction of
Ceiling Tile Systems to Shocks. NIST NCSTAR I-5D. National Institute of Standards and
Technology. Gaithersburg,
MD,
September.
Hamins. A.. A. Maranghides, K. B. McGrattan, T.
J.
Ohlemiller, and R. Anleitner. 2005. Federal
Building and Fire Safety Investigation of the World Trade Center Disaster: Experiments and
Modeling of Multiple Workstations Burning in a Compartment. NIST NCSTAR 1-5E. National
Institute of Standards and Technology. Gaithersburg, MD, September.
McGrattan, K.
B., C. Bouldin,
Investigation of the
and G. Forney. 2005. Federal Building and Fire Safety
World Trade Center Disaster: Computer Simulation of the Fires in the World
NIST NCSTAR 1-5F. National Institute of Standards and Technology.
Trade Center Towers.
Gaithersburg,
MD,
September.
Prasad, K. R.. and H. R.
Baum. 2005. Federal Building and Fire Safety
Investigation of the
World
Trade Center Disaster: Fire Structure Interface and Thermal Response of the World Trade Center
NIST
Towers.
MD,
NCSTAR
1-5G. National Institute of Standards and Technology. Gaithersburg,
September.
and T. McAllister. 2005. Federal Building and Fire Safety Investigation of the World Trade
Center Disaster: Structural Fire Response and Probable Collapse Sequence of the World Trade Center
Gross,
J.
Towers.
L.,
NIST NCSTAR
1
-6.
MD,
National Institute of Standards and Technology. Gaithersburg,
September.
M. A. Stames, J. L. Gross, J. C. Yang, S. Kukuck, K. R. Prasad, and R. W. Bukowski.
2005. Federal Building, and Fire Safety Investigation of the World Trade Center Disaster: Passive
Carino, N.
J.,
NIST NCSTAR
Fire Protection.
Gaithersburg,
Gross,
J.,
F.
MD,
1-6A. National Institute of Standards and Technology.
September.
Hervey, M. Izydorek,
J.
Mammoser, and
J.
Treadway. 2005. Federal Building and
Fire Safety Investigation of the World Trade Center Disaster: Fire Resistance Tests of Floor Truss
Systems. NIST NCSTAR 1-6B. National Institute of Standards and Technology. Gaithersburg,
MD,
September.
Zarghamee, M.
S., S.
M. Mudlock, W.
NIST NCSTAR
1-3D,
I.
WTC
Bolourchi, D.
Naguib, R.
Investigation
W.
Eggers, O. O. Erbay, F.
P. Ojdrovic,
A. T. Sarawit, P.
R
W. Kan,
Barrett,
J.
Y. Kitane, A. A. Liepins,
L. Gross,
and
xxvn
Preface
T. P. McAllister.
2005. Federal Building and Fire Safety Investigation of the World Trade Center
Disaster: Component, Connection,
and Subsystem
Structural Analysis.
National Institute of Standards and Technology. Gaithersburg,
Zarghamee, M.
and Fire
Building
Y. Kitane, O. O. Erbay, T.
S.,
P. McAllister,
MD,
and
J.
L. Gross.
McAlHster,
W. M.
Pitts,
2005. Federal
Damage and Fire. NIST
1-6D. National Institute of Standards and Technology. Gaithersburg,
R.
T.,
1-6C.
Safety Investigation of the World Trade Center Disaster: Global Structural
Analysis of the Response of the World Trade Center Towers to Impact
NCSTAR
NIST NCSTAR
September.
W. Bukowski,
K. R. Prasad,
F.
R. G. Gann,
J.
L. Gross, K. B.
MD,
September.
McGrattan, H. E. Nelson, L. Phan,
Sadek. 2006. Federal Building and Fire Safety Investigation of the World
Trade Center Disaster: Structural Fire Response and Probable Collapse Sequence of World Trade
Center
(Provisional).
7.
Gaithersburg,
NIST NCSTAR
1-6E. National Institute of Standards and Technology.
MD.
Gilsanz, R., V. Arbitrio, C. Anders, D. Chlebus, K. Ezzeldin,
W. Guo,
P.
Moloney, A. Montalva,
Oh, K. Rubenacker. 2006. Federal Building and Fire Safety Investigation of the World Trade
Center Disaster: Structural Analysis of the Response of World Trade Center 7 to Debris Damage
J.
and Fire.
(Provisional).
Gaithersburg,
NIST NCSTAR
and Technology.
1-6F. National Institute of Standards
MD.
Kim, W. 2006. Federal Building and Fire Safety Investigation of the World Trade Center
Seismogram Data. (Provisional). NIST NCSTAR 1-6G.
Disaster: Analysis of September 11, 2001,
National Institute of Standards and Technology. Gaithersburg,
MD.
Nelson, K. 2006. Federal Building and Fire Safety Investigation of the World Trade Center
Disaster: The
Con Ed Substation
in
World Trade Center
7.
(Provisional).
National Institute of Standards and Technology. Gaithersburg,
Averill,
J.
NIST NCSTAR
I-6H.
MD.
D., D. S. Mileti, R. D. Peacock, E. D. Kuligowski, N. Groner, G. Proulx, P. A.
Reneke, and
H. E. Nelson. 2005. Federal Building and Fire Safety Investigation of the World Trade Center Disaster:
Occupant Behavior, Egress, and Emergency Communication. NIST
Standards and Technology. Gaithersburg,
MD,
NCSTAR
1-7.
National Institute of
September.
Fahy, R., and G. Proulx. 2005. Federal Building and Fire Safety Investigation of the World Trade
Center Disaster: Analysis of Published Accounts of the World Trade Center Evacuation.
NCSTAR
Zmud,
J.
1-7A. National Institute of Standards and Technology. Gaithersburg,
Lawson,
J.
R.,
of Standards and Technology. Gaithersburg,
and R.
L. Vettori. 2005.
MD,
NIST NCSTAR
1-7B. National
September.
Federal Building and Fire Safety Investigation of the World
Trade Center Disaster: The Emergency Response Operations.
Standards and Technology. Gaithersburg,
xxvni
September.
2005. Federal Building and Fire Safety Investigation of the World Trade Center
Disaster: Technical Documentation for Sun'ey Administi-ation.
Institute
MD,
NIST
MD,
NIST
NCSTAR
1-8.
National Institute of
September.
NIST NCSTAR
1-3D,
WTC Investigation
Acknowledgments
Many people
contributed valuable
skills, facilities
and information
that
made
this report possible.
We
thank:
Professor George Krauss (University Emeritus Professor, Colorado School of Mines) and Dr. Elliot
Brown (EB
the
Consulting. Golden,
CO)
for technical discussions regarding the metallurgy
of the
steels in
World Trade Center,
Dr Hugh MacGillivray
(Professor, Imperial College,
London,
UK)
for technical discussions regarding
high strain rate testing.
Professor David Matlock (Colorado School of Mines) for use of the high-rate tensile testing machine,
Tessa Dvorak and Brian
Goudy
(National Institute of Standards and Technology [NIST]) for data
reduction,
Lonn Rodine (NIST)
for
specimen and drawing preparation,
Ray Santoyo (NIST) and Mark
ladicola (NIST) for conducting
many of the room temperature
tensile
tests.
Donald
Hame (NIST
)
for conducting
some of the creep
tests,
Richard Rhorer and Michael Kennedy (NIST) and Debasis Basek (University of Tennessee) for
conducting the high-rate Kolsky bar
Dr. T. Weerasooriya and Mr. P.
Directorate,
assisting in
tests.
Moy
(Impact Physics and High Rate Laboratory,
Army Research Laboratory, Aberdeen, MD)
calibrating the NIST Kolsky bar facility,
for performing initial
Richard Ricker and David Dayan (NIST) for conducting the
tests to
Weapons and
Kolsky bar
tests
Materials
and
determine the elastic modulus as a
function of temperature,
Kazushige Tokuno (Nippon Steel
Former employees of Laclede
locating Laclede records for
NIST NC STAR
1-3D,
WTC
USA)
for supplying historical information
Steel (St. Louis,
NIST
MO)
including David
on Yawata
McGee and
steel properties.
Larry Hutchison for
investigators during a visit to the Laclede Steel Mill.
Investigation
XXIX
Acknowledgments
This page intentionally
XXX
left
blank.
NISTNCSTAR
1-3D,
WTC Investigation
Executive Summary
National Institute of Standards and Technology (NIST) had several objectives in characterizing the
recovered World Trade Center
(WTC)
recovered from the impact and
fire
and deformation
rate to
to.
One was
•
Elastic properties as a function
modeling,
steels
from the construction
that they
era.
data can be divided into five groups:
of temperamre were determined from recovered perimeter
steels.
Room temperature yield
specimens of recovered
and
tensile strength,
steel
and
total
of all grades from the
strain cur\'es are reported for
29 different
strengths of several different perimeter
Yield and tensile strengths and
were determined
total
elongation were determined from
fire
and impact zones. Complete
steels with 12 yield strength levels.
A
displacement curves were measured from recovered
•
fire
and welds were consistent with both the specifications
steels, bolts
and the expected properties of structural
column
mechanical properties of steels
zone as a function of temperature to provide data for
The experimentally determined mechanical property
•
to characterize the
provide data for impact modeling. The second was to determine whether the
mechanical properties of the
were delivered
steels.
column and
325
truss
stress-
Load-
bolts. Stress-strain curves
weld geometries
and
are reported.
elongations of selected perimeter and core column steels
as a function strain rate at rates
up
to
400
s"'.
Strain-rate sensitivities
and
complete stress-strain curv es for these steels are reported. Several steels were characterized
higher rates using Kolsky bar
•
at
tests.
Impact properties as a function of temperature were determined using Charpy
tests for selected
perimeter column, core column, truss, and truss-seat steels.
•
Elevated-temperature yield and tensile strength, and total elongation were determined on
selected specimens of recovered perimeter and core column, truss, and truss seat steels.
Complete
stress-strain curves for temperatures
up
to
650 °C are reported
characterized. Creep deformation as a function of temperature
and
stress
for all steels
was determined
for
the truss chord steel.
By combining
these measured properties with historical averages from the literature, and in
recovered mill
test reports,
NIST developed
some cases
values for properties of the important grades of steel
including
modulus and Poisson's
•
Elastic
•
Model room-temperature
ratio as a function
of temperature,
stress-strain curves for all
WTC
steel
grades from the
fire
and
impact zones corrected for dynamic effects,
•
Room-temperature load-displacement curves for
NISTNCSTAR
1-3D,
WTC Investigation
A 325 bolts.
XXXI
Executive
Summary
•
Yield and tensile strength for weld metals,
•
Strain rate sensitivity of steel for high strain rate properties,
•
Elevated-temperature yield and tensile strength and complete stress-strain behavior for
WTC
steel
grades from the
fire
and impact zones,
•
Elevated-temperature load-displacement curves for
•
Creep response for
all
WTC
steel
grades from the
A 325 bolts,
fire
and impact zones.
This report makes several findings concerning the steel used in the
•
Steels, bolts,
all
and welds generally have properties
WTC:
that are consistent with the specifications to
which they were supplied.
•
Infrequently, the measured yield strengths are lower than called for
specifications.
The measured
yield strengths of 2 of 24 perimeter
by the appropriate
column samples
are lower
than the specified minimum. This probably occurred from a combination of the natural
variability of steel properties,
The measured
protocols.
specified
minimum.
and small differences between the mill and NIST testing
yield strengths of 2 of 8 core
In this case, mechanical
damage
column samples
are lower than the
that occurred in the collapse
and
subsequent recovery removed the expected yield point behavior and lead to the low values.
•
The average measured
specified
minimum
yield strength of the steels
from the perimeter columns exceeds the
values by about 10 percent, which
is
consistent with historically expected
values for steel plates.
•
The measured
yield strengths of the F,
= 36
ksi wide-flange core
expected from historical measurements of other structural
•
The
strain rate sensitivities
The impact
from the
column
•
XXXll
from the
WTC construction era.
properties of the steels, evaluated
by Charpy
testing, are similar to structural steels
WTC era. The ductile-to-brittle transition temperatures of the perimeter and core
steels are at or
The behavior of the
that
steels.
of the yield and tensile strengths of perimeter and core columns are
similar to other structural steels
•
columns are lower than
below room temperature.
WTC steels with temperature is similar to
WTC construction era.
yield and tensile strengths of
of other structural
steels
from the
NISTNCSTAR
1-3D,
WTC Investigation
Chapter
1
Introduction
OVERVIEW OF REPORT
1 .1
National Institute of Standards and Technology (NIST) had three objectives in characterizing the
mechanical properties of the structural
(WTC)
site.
The
first
was
specifications that they
to
steels, bolts
and welds recovered from the World Trade Center
compare the measured properties of the
were purchased under. The second was
consistent with the properties and quality of other structural steel
third
was
to determine the constitutive behavior
with the requirements of the
steels
to determine
from the
of the steels for input into
whether
their properties
were
WTC construction era. The
element models of the
finite
response of the building components to the airplane impact and to the high temperatures produced by the
subsequent
fires.
In support of these three goals, the Investigation characterized the elastic properties at
elevated temperatures, the room-temperature tensile behavior, the high-strain-rate behavior, the impact
behavior using Charpy
tests,
and the elevated-temperature tensile and creep behavior.
This report comprises five chapters that summarize the results of investigations into the mechanical
properties of the
Chapter
WTC
3 co\ ers the
steels.
Chapter 2 covers the temperature dependence of the elastic properties.
room-temperature, quasi-static stress-strain behavior. Chapter 4 details
investigations into the properties of the steel at high strain-rate,
airplane impact. Chapter 5 summarizes the results of
Charpy
which
are relevant for
tests to characterize the
modeling the
impact properties of
the steels. Chapter 6 addresses the elevated-temperature deformation behavior of the steel,
which
is
where appropriate,
relevant for modeling the response of the steel to the fires. Within each chapter,
individual sections address the relation of the measured properties to the standards and specifications used
in the
WTC
WTC
construction era as a
construction.
Each chapter also compares the measured properties
means
to establish
whether the properties of the
ordinary. Finally, separate sections explain the
to literature data
steel
methods by which experimental
from the
were anomalous or
results, literature
and
construction data, and theoretical models were combined to produce constitutive laws for input to the
finite
1.1.1
element models.
Elastic Properties (Chapter 2)
After examining reported literature values of the change in elastic properties with temperature, the
Investigation characterized specimens of perimeter
and probably
results
from differences
column
in test technique.
steels.
The
scatter in the literature data
is
large
Analysis of the experimental data generated in the
Young's modulus, shear modulus, and Poisson's ratio for
700 °C. For higher temperatures, the Investigation produced a methodology for
Investigation produced expressions for
temperatures up to
estimating the modulus.
1.1.2
Room Temperature Tensile
Properties (Chapter 3)
Metallurgically and mechanically, only the room-temperature stress-strain behavior detailed in Chapter 3
is
relevant to the standards and specifications that the steel
NIST NCSTAR
1 -3D,
WTC
Investigation
was
to meet. In support
of the
first
objective,
1
Chapter
NIST
1
characterized the room-temperature tensile stress-strain behavior of several hundred tensile
specimens from the perimeter columns, core columns,
samples of all major strength levels of steel from
impact and
fire
zones of the towers. In addition,
trusses,
and
the fabricators
all
NIST
The specimens represent
truss seats.
who
supplied steel for the floors in the
characterized the load-displacement behavior of
recovered bolts and the strength of selected welds and weld assemblies.
In
comparing the properties of the recovered
steels to their intended specifications,
analyses, room-temperature tensile tests, and archival information.
complicated the positive identification of the exact specifications used for a given
limitations, the Investigation
made
NIST used chemical
A number of experimental
difficulties
Despite these
steel.
several findings regarding the properties of the recovered steel in
relation to the original specifications.
The
yield and tensile strengths of the perimeter columns, with only a
the strength requirements of the original specifications.
the
amount by which they
fall
short
is
The number of slightly under-strength
plates
and
consistent with expected values for the average strength and
coefficient of variation of yield strengths of plate steels from the
measured
few exceptions, are consistent with
yield strength to specified yield strength
is
WTC construction era. The ratio of
also consistent with literature estimates from the
WTC construction era.
Unlike the perimeter columns, the NIST-measured yield strengths of several of the wide-flange core
columns recovered from the core were
damage
3 ksi
that
removed
below the value
less than specified.
which could
the yield point behavior,
in the mill test report.
It is
likely that these
easily lead to
The average measured
Many
expected value predicted from the
of the components
in the floor trusses
reports
ASTM A 325
from the contemporaneous
The
limited
up
to
minimum
are
still
several ksi
WTC construction era, however.
A 36 minimum requirements.
specification, but are
literature.
strengths
were fabricated from high-strength, low-alloy (HSLA)
even when they were only required to meet the
bolts are consistent with the
of the
literature
NIST-measured
yield strengths and yield points for the
wide-flange specimens whose yield points were larger than the specified
less than the
low values arose from
much
The
steels,
strengths of recovered
higher than expected based on
number of tests on welds
indicates that their
strengths are consistent with the expected values from the welding procedures used.
To provide
constitutive data for finite element
models of the Investigation, the
third objective for the
room-temperature tensile testing program, NIST produced representative stress-strain curves for each of
the
many
grades of steel in the towers. For the perimeter column and truss steels, the yield strengths were
corrected for testing-rate effects and, where possible, experimental data were combined with surviving
WTC mill test reports to produce better estimates of the characteristic strength.
strain curves
NIST-measured
stress-
from specimens taken from recovered components, modeled using the Voce work-hardening
law, were used to describe the plastic deformation of each steel grade. For the core columns, either plates
or wide-flange shapes, the yield strength
the
was assumed
to
WTC construction era, corrected for dynamic effects.
be the historical average from the literature of
NIST-measured
stress-strain curves
from
specimens recovered from core columns, also modeled using Voce work hardening, supplied the
plastic
behavior. For the truss steels, based on chemistry and mechanical characterization, the angles were
assumed
to
were used
be a high-strength low-alloy
to estimate the yield
steel regardless
displacement curves measured on recovered
2
of specification. Experimental stress-strain data
and work-hardening behavior. For the
bolts.
A
325
NIST supphed loadcolumn welds, NIST
bolts,
For the perimeter and core
NISTNCSTAR
1-3D,
WTC Investigation
Introduction
estimated mechanical properties by combining data from archival materials and tests on welds in
recox ered components.
High-Strain-Rate Properties (Chapter 4)
1.1.3
Because the strength of steel depends on the
rate at
which
it is
deformed,
it is
necessary to quantify this
relation for the steels in the impact zone. Failing to properly account for the increased strength with
deformation rate could lead to an incorrect estimates of the damage caused by the airplane impact. In
support of this goal, the Investigation employed two types of high-strain-rate
50
s~'
<s < 500 s~'
tests.
For
rates,
tests used a serv'ohydrauhc tensile test machine and special tensile specimens. For
,
higher rates, tests employed a Kolsky bar
test
apparatus in which the test specimen
is
a right circular
cylinder loaded in compression.
Like other structural
steels, the strength
WTC
of the
rate sensitivity
steels
of the
WTC steels increase with increasing strain rate. The strain
was evaluated
for eight perimeter
columns, and one plate from a core box column. The measured
magnitude as other structural
steels reported in the literature
Impact Properties (Chapter
1.1.4
Charpy impact
tests are a type
before fracturing.
As
plates, four wide-flange core
5)
of dynamic fracture
is
test that
probes the ability of steels to absorb energy
measured
Charpy
test,
the
as a function of temperature.
ASTM International (ASTM) specifications for steels used in the WTC, then or current, put
limits
on the impact properties, but the measured impact properties are similar
steels
of the
WTC
to those
of other structural
construction era. All the perimeter column steels have large upper shelf energies and
transition temperatures well
truss rods
of the same
and decrease with increasing yield strength.
such, they are particularly relevant to the airplane impact. In the
energy used to break a notched specimen
None of the
column
strain rate sensitivities are
below 0
°F.
The
transition temperatures
of the wide-flange specimen and the
and angles are near room temperature. The transition temperatures of the truss-seat
above room temperature, and the absorbed energy of these
steels is
low even
at
room
steels are
temperature,
indicating a propensity for brittle failure.
Elevated-Temperature Properties (Chapter
1.1.5
The high-temperature
stress-strain
was
testing
program had two
One was
to characterize the elevated-temperature
behavior of the steels most likely to have been affected by the post-impact
to characterize the creep, or
trusses. In
thrusts.
6)
each of these two areas,
methodologies
to predict the
fires.
The second
time-dependent deformation, behavior of the steels from the floor
in addition to the
behavior of untested
experimental characterization,
NIST developed
steels.
For the elevated-temperature stress-strain behavior, the methodology recognized that the yield and tensile
room-temperamre values, follow a master curve with
master strength curve, developed using literature data on bolt steels,
strengths of structural steels, normalized to their
temperature.
A
describes the
more
NISTNCSTAR
modified form of this
1-3D,
rapid strength degradation of the bolts with temperature.
WTC
Investigation
3
Chapter
1
To produce elevated-temperature
stress-strain curves
of WTC
NIST developed
steel,
a
methodology
to
account for the change in the work-hardening behavior using literature data for structural steels scaled by
the ratio of room-temperature tensile strengths.
NIST
also characterized the creep deformation behavior of the floor truss steels.
NIST developed
properties of untested steels,
a
methodology
that
To
estimate the creep
used either existing
literature data or the
Investigation-generated floor truss creep data after scaling by the ratios of room-temperature tensile
strength.
DESCRIPTION OF THE MAJOR BUILDING COMPONENTS
1.2
This section suminarizes the major structural elements of the
NCSTAR
(NIST
The
WTC buildings
in the
impact and
fire areas.
1-3A') contains a more detailed description of the structure of the building.
WTC towers had a frame-tube construction consisting of closely spaced perimeter columns with a
rectangular core.
The buildings had
a square footprint,
the 9th to the 107th floors, the perimeter
the building
system
that
was approximately 87
ft
207
ft
2
in.
on a side with chamfered comers. From
columns were closely spaced, built-up box columns. The core of
by 137
ft
and connected
to the perimeter
columns by a floor panel
provided column-free office space.
Perimeter Columns
1.2.1
Outer
14.0
in.
Web
(Plate 2)
/
Each building face consisted of
59 columns spaced
coluiTuis
at
40
in.
The
Flange
(Plate 1)
were fabricated by welding
13.5
in.
Section at individual column
four plates to form an approximately
14
in.
square section. Fig. 1-1. The
15.75
perimeter columns were joined by
horizontal spandrel plates to
panels,
in
inner
Web
form
Outside of building
which were typically three
40.0
Heavy
wide. Fig. 1-2.
in.
and three columns
stories (36 ft) tall
end, or "butt"
plates 1.375 in. to 3 in. thick
welded
(Plate 3)
Splice Plates
were
and bottom of each
to the top
column. The columns were also bolted
Inside of building
to the adjacent
A
columns using
ASTM
Spandrel Plate (Plate 4)
325 bolts except for the heaviest butt
plates,
which used
Other than
at the
ASTM A 490 bohs.
mechanical
floors, the
panels were staggered vertically so that
only one third of them were spliced in any
one
story.
Section at spandrel
Figure 1-1. Cross section of a perimeter column;
sections with and without spandrels.
Adjacent spandrels were bolted
together using splice plates.
This reference
is
to
one of the companion documents from
this Investigation.
A hst of these documents
appears in the Preface
to this report.
4
NIST NCSTAR
1-3D,
WTC Investigation
Introduction
The
structural steel
minimum
documents specified 14 grades of steel
for the plates
of the perimeter columns, with
yield strengths of (36. 42. 45, 46, 50, 55, 60, 65, 70, 75, 80, 85, 90, and 100) ksi.
Source: Unknown. Photo enhanced by NIST.
Figure 1-2. Characteristic perimeter column panel illustrating the various components.
Designations in parentheses refer to the specimen nomenclature of Table 1-1.
The
side plates of a
column
are termed the flanges, Fig. 1-1, and the inner and outer plates are termed the
webs. In an individual column, the flanges were always
at least as thick as the
strengths of the
webs and flanges of an individual column were
columns within
a three-story, three-column panel could differ.
As
the elevation in the building increased, the thickness of the plates in the
plates
were always
at least
0.25
in. thick.
webs. Although the yield
identical, the yield strengths
columns decreased, but the
About one half of the recovered perimeter columns
in the
web plates are both 0.25 in. thick. In the WTC
type make up about 2/3 of the total columns
inventory are type 120, in which the flange and
impact zone, floors 92-100, columns of this
of the
1
NIST
fire
and
Twelve grades of steel were specified for the spandrels, with the same strength levels as the columns but
with a maximum F, ^ 85 ksi. In a panel, the yield strength of the spandrel plate was generally lower than
that
of the webs and flanges, and
its
spandrels crossed the columns there
NISTNCSTAR
1-3D,
WTC
Investigation
was greater than
was no inner web plate.
thickness
the adjacent inner
web
plate.
Where
the
5
Chapter
1
Core Columns
1.2.2
Core columns were of two types: welded box columns and rolled wide flange (WF) shapes.
shows some examples of the shapes of core columns.
very large box columns, as large as 12 in.x52
In the lower floors, the core
with plates up to 7
in.
in. thick.
Fig. 1-3
columns were primarily
In the upper floors, the
columns were primarily rolled wide-flange shapes. Like the perimeter columns, the core columns were
typically spliced at three-story intervals.
Core box columns were specified with F, = 36
Core wide-flange columns were specified
Fy = 42
to
ksi or
Fy = 42
be one of four grades, but were primarily F, = 36
ksi
ksi.
and
ksi steel.
22.0
I
in.
J
I
^
J
I
i
1.31
.
,
in.
t
14.0
in.
1
1
t
Column 504
22.0
in
Column
15.0
X-
1
\
s
Columns 1001 1008
501 508
In.
701
1
1
Columns 607 906
14WF219
iD.64
17.89
15.825
in.
10.00
in.
in.
in.
Note: To scale.
Figure 1-3. Typical welded box columns and rolled wide-flange shapes used for core
columns between the 83rd and 86th floors.
Flooring System
1.2.3
Other than
in a
few
special mechanical floors, the floor outside the core
dimensional network of 29
individual floor truss.
in.
deep floor trusses. Figure 1-4
The most common
floor trusses
was supported by
illustrates the
were nominally 60
major components of an
ft
or 36
ft
were dozens of variants, the top chord was usually fabricated from two sections of 2
thick angle specified to conform to
was usually
fabricated from
A 36.
The 36
truss
web was
D
6
= 0.92
in.
ft
two
ASTM A 242 with F,
= 50
ksi. In the
a two-
60
ft
long.
Although there
in. x 1.5
in.xO.25
slightly larger angles, 3 in.x2 in.xO.37 in. thick specified to
conform
trusses generally used a 2 in.xl.5 in.x0.25 in. thick angle for the lower chord as well.
a continuous round bar,
otD =
0.98
in.
for the
36
which was usually D= 1.09
ft
in.
for the
in.
trusses, the lower chord
60
ft
trusses
to
The
and
trusses.
NISTNCSTAR
1-3D,
WTC Investigation
Introduction
Gusset f^ate
Core Column
Figure 1-4. Schematic diagram of a floor truss.
Before erection, the individual trusses were assembled into modules typically 20
ft
wide. Each module
contained two sets of doubled trusses in the interior and a single truss along each edge. Smaller bridging
main
trusses ran perpendicular to the
fabricated from 1.5
At the
in.
x 1.25
core, the floor trusses
in.
trusses at 13
x 0.23
in.
were bolted
ft
4
angles and
to seats
spacing.
in.
D
=
0.75
on channels
in.
The bridging
trusses
were usually
round webs.
that ran continuously along the core
columns. At the perimeter columns, the floor trusses were bolted and welded to seats mounted on the
spandrels at every other column. Fig. 1-2. Diagonal bracing straps welded to gusset plates secured the
installed floor truss
1.3
modules
to the perimeter
columns. Fig. 1-2.
SPECIMEN NOMENCLATURE
The specimen nomenclature of the Investigation captures
the vital information about the location of the
individual specimens within the larger component. Because the recovered sections arrived at
unidentified, and
it
often required
weeks or months
NIST
to establish the original location in the building, the
NIST report "Steel Inventory and
tables that map the specimen
nomenclature does not capture details about the original location. The
Identification within Buildings"
designation to
its
NCSTAR
1-3B) contains
location in the building and specified dimensions and strength. Furthermore, because of
the significant damage,
NIST NCSTAR
(NIST
1-3D,
it
WTC
was not possible
Investigation
to identify all the
recovered sections.
1
Chapter
1
Table 1-1 describes the naming convention for the perimeter column samples with an example.
Figure 1-2 illustrates the naming convention for the perimeter column and truss components graphically.
The example specimen name, N8-C1T2-IW-L2, describes the second specimen in a series of longitudinal
tension specimens from the inner web of a perimeter column panel from WTC 1 column line 143 near top
of the panel
at floor 100.
impact zone. In
scheme used
this
This panel
is
from the north face of the tower,
above and
nomenclature, the column numbering within a given panel
in the structural
all
1
of the
reversed from the
other specimens using an example from a core column.
The example specimen name, C65-F1, describes the
WTC
is
to the east
engineering drawings.
Table 1-2 describes the naming convention for
flange column from
just
on column
line
first
sample taken from the flange of the core wide-
904 somewhere between floors 86-89. This column
and on the opposite side of the building from the impact zone. The sample
that the
is
below
example describes
supplied the tension specimens used for characterizing the tensile behavior of the core columns.
1.4
SYMBOLS AND ABBREVIATIONS
The preface
to this report contains tables
of acronyms and units used
in this report.
Table 1-3 defines
several symbols and ternis relating to mechanical testing.
NISTNCSTAR
1-3D,
WTC Investigation
Introduction
Table 1-1. Specimen nomenclature for perimeter column specimens.
Example N8-C1T2-1W-L:
Sub
Element
Possible
Values
Description
N8
Reco\ ered item tag
CI
Describes column # in panel''
CI
Column
1
C2
Column
2:
center
C3
Column
3:
column on
T
:
column on
the
denotes the recycling yard where specimen originated^
left,
T
Near top
M
Near middle
B
Near bottom
1,
IW
looking into the building
column
the right looking into the building
Describes specimen location
-»
2. 3 etc., chronological
in
column
by specimen cut from
this larger section
Describes plate location within column or other component within assembly
IW
Inner
LF
Left flange (looking into building); sometimes abbreviated as
RF
Right flange (looking into building); sometimes abbreviated as
OW
Outer w eb
w eb'
S
Can
refer to either spandrels or truss seats
SP
Can
refer to either spandrels or splice plates
Step
EP
G
or butt plate of
L
Plate
that ties floor truss to
welded
to
L
Longitudinal
T
Transverse
1.2,3...
column
bracing strap to perimeter column panel
that attaches diagonal
perimeter column panel and truss
Describes orientation with respect
2
FR
column
Diagonal bracing strap
Gusset plate
Tab
FL
Truss seat
End
DBS
a.
(letter
Specimen number w ithin
to rolling direction''
series of orientations
engineenng drau ings identify the perimeter column panels by the column number of the center column
142 97-100 is tower I, column line 142 floors 97-100. In the
For example, WTC
Investigation notation, this is recovered panel N8. The structural engineering drawings number the columns looking from
The
strucniral
and the floors
at the splices.
1
outside the building.
b.
CI corresponds
example above.
c.
The
to
column
line 143;
C2 corresponds
structural engineering plans also
number
to
column
For spandrel
Therefore
all
NIST NCSTAR
later
to
RF
S or SP corresponds to plate
plates, the orientation designation refers to the orientation
Metallographic analysis
C3 corresponds
the four plates in each perimeter column:
OW corresponds to plate 2, IW corresponds to plate 3. and
d.
line 142;
column
and
LF
line 141 in the
correspond to plate
1,
4.
with respect to the vertical axis of the building.
revealed that the rolling direction of the spandrel plates
is
along their long direction.
spandrel specimens designated with an "L" are oriented transverse to the rolling direction.
1-3D,
WTC
Investigation
9
Chapter
1
Table 1-2. Specimen nomenclature for core box and wide-flange shapes, trusses, and
other specimens.
al
Example C65-F1
SUD
Possible
Element
Values
Description
C65
Recovered item tag (denotes the recycling yard where the specimen was recovered)
F
Describes location within recovered item
WF shapes or wider plate for box columns)
Flange (for WF shapes or wider plate for box columns)
Web (for WF shapes or narrower plate for box columns)
Web (for WF shapes or narrower plate for box columns)
F
Flange (for
FL
W
WEB
BA
Bottom chord bulb angle
TA
Top chord bulb angle
TR
Round bar
LR
Large diameter round bar
MR
Medium
SR
Small diameter round bar (for truss webs)
TS
Core
Weld
(for truss
(for trusses)
(for trusses)
webs)
(for truss
webs)
diameter round bar (for truss webs)
truss seat
Taken from
a
weld
component
in the
Describes orientation with respect to rolling direction (used for tensile specnnens)
L
Longitudinal
T
Transverse
1, 2, 3...
1
Specimen number within
series
of orientations
Table 1-3. Mechanical testing definitions used
in this report.
Yield strength
Generally used as the specified
use.
ROA
reduction of area
TS
Tensile Strength
(ASTM
usage Standard
The maximum engineering
Often referred
YP
minimum
E
6)
stress in a tensile test
(ASTM E
6 usage)
to in other literature as ultimate tensile strength,
UTS. or by the symbol
5„.
Yield point
The
first
maximum
engineering stress-engineering strain curve, associated with discontinuous
in the
yielding. Also called upper yield strength
This report uses the symbol
minimum
YS
yield strength or yield point depending on the steel standard in
(AISC usage)
YP
(ASTM E
to denote a
6 usage).
measured value of the yield point as opposed
to
its
Yield strength
The engineering
uses the 0.2
stress at
which
it
is
considered that plastic elongation has commenced. This document
% offset stress. (ASTM E 6 usage)
This report uses the symbol YS to denote a measured value of the yield strength as opposed to
specified
EI,
minimum
its
value.
Total Elongation
The elongation (of a
tensile
specimen) determined
broken ends of the specimen.
10
specified
value.
(ASTM E
after fracture
by realigning and
fitting
together of the
6 usage)
NISTNCSTAR
1-3D,
WTC Investigation
Chapter 2
Elastic Properties
INTRODUCTION
2.1
This section repons the expressions for the temperature dependence of the elastic properties of steel
Team members. Expressions include Young's modulus £, Poisson's
ratio V, and shear modulus G. World Trade Center (WTC) structural steels can be assumed to be
elastically isotropic, so the shear modulus is related to E and v:
supplied to other
NIST
Investigation
+ K)
2(1
Because
structural steels contain relatively
difference in
modulus between
From room temperature up
structural steel
is ferrite,
to
low
in
pure iron, but lower
is
no
significant
about 720 °C the primary metal phase in iron-carbon alloys such as
usually denoted as a-phase.
y-gamma
transformation to a mixture of a-phase and
910 °C
fractions of alloying elements, there
structural steel variants.
in
most
Above
this
temperature steel undergoes a phase
phase, or austenite. At
still
higher temperatures,
structural steels, the phase transformation to all austenite
complete. In the two-phase region, the fraction of a-phase and
y-gamma
is
phase, and therefore modulus,
depends strongly on carbon content and temperature. This phase change introduces some difficulty
in
representing the change in modulus with temperature.
EXPERIMENTAL PROCEDURE
2.2
Three specimens of WTC
analyzer (Model
steels,
summarized
in
Table 2-1, were characterized
in a
thermomechanical
DMA 2980, TA Instruments) over the temperature range -140 °C < T<600 °C.
mmxlO mmxl mm were loaded in three-point bending. The
Rectangular specimens nominally 50
thermomechanical analyzer oscillated the specimens
Measurements were conducted
second was
in the
in
range 50 °C <
two
stages.
r< 600
The
hz
at a
was
at
temperatures,
The
chamber was
ELASTIC PROPERTIES
2.3
literature is rich
filled
r<
(E, v,
^m.
50 °C. The
made
at 5
°C
During the high-temperature
with an inert gas to suppress oxidation.
G)
FOR
0°C < T < 723 °C
with determinations of the change of Young's modulus, E, with temperature for
r<723 °C. Figure 2-1 shows some examples. The wide
different
-140 °C <
°C. In both cases, modulus determinations were
intervals while the furnace temperature increased at 5 °C/min.
characterization the specimen
constant amplitude of 15
at 1.0
first
measurement techniques. The data
variation in the values
sets with the highest values
is
an
artifact
of the
were obtained using ultrasonic
techniques. These data represent adiabatic determinations, and are not appropriate for estimating elevated-
temperature modulus
NISTNCSTAR
1-3D.
in structures.
WTC
The data
Investigation
sets with the lowest values
were probably determined from
11
Chapter 2
the loading portion of high-temperature tensile tests, and therefore include inelastic specimen
deformation. These determinations are also inappropriate.
Figure 2-1 also plots Young's modulus, E, determined for the three specimens of perimeter columns,
summarized
in Tab. 2-1,
determined using the thermomechanical analyzer.
A third-order polynomial was
summarized
in
sufficient to represent the data for 0
°C < r<600 °C
for the three steels
Table 2-1 and plotted in Fig. 2-2.
EiT) =
e^^
+
ej + ej^'
+ e^T^
where
eo= 206.0
GPa
e,=
-0.04326 GPa/°C
62=
-3.502xlO-^GPa/(°C)-6.592x10-^ GPa/(°C)-
The constant term,
polynomial
data that
fit.
was
eo,
fixed as the average of the three steels at 0 °C and
The room-temperature value
for
was not
part of the
E for the fit lies within the range of the
highly regarded
Galambos (1978) recommends.
Figure 2-3 summarizes the relatively scarce literature data for Poisson's
temperature.
significantly.
The
The
ratio, v, as a
function of
Cooke 1988) do not differ
data for 0 °C < T < 725 °C:
data for iron (Dever 1972) and mild steel (Clark 1953 cited in
solid line is the
fit
of a fourth-order polynomial to
v{T) =
n^^
+
riyT
all
+ n^T' + n^T^ +
the
n^T'^
where
^
770=
0.28737362
/?,=
2.5302417x10"-
n2=
773=
2-3
(°C)"'
2.6333384x10"^ (°C)"-
-9.9419588x10 "
{°C)-^
1.2617779x10"'-' (°C)"^
for
the
T in the range 0 °C < T < 725 °C. The room-temperature value, (v = 0.288), is 4
commonly used value v = 0.3, which has been rounded to one significant digit.
The shear modulus, G, was determined from
Fig. 2-2, values for the shear
modulus
E and
using Eq. 2-1
as a function of temperature,
points from Eq. 2-1 using the expressions for
fit
v,
E (T) and v
.
To
percent smaller than
generate the curve
G (T),
were calculated
shown
in
at discrete
(7), Eqs. 2-2 and 2-3. The resulting data were
using a fifth-order polynomial.
G{T) = g, +g,T +
gj' +g,T'+gX +gsT'
where
80.005922 GPa
go=
12
gi=
g2=
-0.018303811 GPa/°C
-1.5650288x10"^ GPa/(°C)'
g3=
-1.5160921x10
g4=
-1.6242911x10"" GPa/CX)"*
g5=
7.7277543x10"'-' GPa/(°C)^
'
2-4
GPa/(°C)-'
NISTNCSTAR
1-3D,
WTC
Investigation
Elastic Properties
for Tin the range 0 < 7 < 725 °C. The room-temperature value of the shear modulus, G, calculated this
way differs by 3.5 percent from the actual measurements cited in Galambos (1978) and the Metals
Handbook (ASM 1978), but it is consistent with the values for E and v from Equations 2-2 and 2-3.
In using the expressions for
stress
of structural
steel
modulus
at
elevated temperature,
it is
important to recognize that the flow
may be
above the a-y phase transformation
only
1
0 percent or less of the room-
temperature value. Calculations that use the high-temperature value of the modulus and do not account for
the very reduced strength
inaccurate.
ELASTIC PROPERTIES
2.4
The
may be
literature data for elastic
(E, v,
FOR r>
G)
910 °C
modulus of y-Fe, the high-temperature phase,
is
sparse. Koster (1948)
provided data for a function for calculating the modulus, Ey(T), of austenite (y-phase) in the range
910 °C < T < 1000 °C, which
is
216.0
yi=
also plotted in Fig. 2-1.
GPa
2-5
26.85 °C
Y2=
73=
4.7x10"' °C'
Data for the Poisson's
ratio for austenite are not available.
Eq. 2-1 after a suitable estimate of Poisson's ratio
ELASTIC PROPERTIES
2.5
In principle,
temperature
it
is
two-phase
steel,
field.
because of
Some of these assumptions
•
G)
made.
FOR 723
°C <
T< 910
possible to do a rule-of-mixtures calculation to evaluate the
in the
even then, each
(E, v,
is
The shear modulus, G, can be calculated from
its
Many
°C
modulus
of
as a function
assumptions are necessary to make the problem tractable, and
composition, would have a different value for the composite modulus.
include
The extrapolation of Eq. 2-5
correctly describes the
modulus of austenite
in the
two-phase
field.
extrapolation of Eq. 2-2 correctly describes the
•
The
•
All alloying elements other than carbon have
no
modulus of ferrite
effect
in the
two-phase
field.
on the positions of the phase
boundaries.
•
The phase change
is
infinitely fast so that the
good approximation on heating, but
The
it is
time-dependence can be ignored. This
much
a
less appropriate for cooling.
uncertainties that these assumptions introduce probably exceed the
mixmres
may be
added
fidelity that a rule-of-
calculation might provide. Instead of the complicated rule-of-mixtures calculation, a
simpler method for approximating the elastic modulus in the two-phase region, Ea.y
is
much
to interpolate
between the two endpoints:
NISTNCSTAR
1-3D.
WTC
Investigation
13
Chapter 2
E^U,)-E^{A,)
2-6
A, -A,
where
= 723 °C and
yi,
= 910
°C.
UNCERTAINTIES
2.6
The expression
for
Young's modulus, Eq. 2-2, represents the experimental data obtained
(Table 2-1) to within 2 percent for
all cases, Fig.
at
NIST
2-4
REFERENCES
2.7
AISC. 1973. Manual of Steel Construction. American
ASME.
Institute
2004. Boiler and Pressure Vessel Code. Section
Subpart 2 Physical Properties Tables, Table TM-1.
Clark, C.L. 1953. High-temperature Alloys. Pitman,
Cooke, G.M.E.
1
988.
An
of Steel Construction.
Materials, Part
II
D
New York p.
6-11.
Properties (Customary),
ASME International. New York. p.
696
New York.
Introduction to the Mechanical Properties of Structural Steel at Elevated
Temperatures. Fire Safety Journal. 13 45-54.
Dever, D.J. 1972. Temperature Dependence of the Elastic Constants in a-Iron Single Crystals:
Relationship to Spin Order and Diffusion Anomalies.
Fields, R.J. T. Weerasooriya,
Stainless Steels, and
One
and M. F. Ashby.
1
J.
Appl. Phys. 43
Koster,
3293-3301.
980. Fracture-Mechanisms in Pure Iron,
Ferritic Steel. Met. Trans. A.
Two
Austenitic
Div.
ASCE. 104
UA 333-347.
Galambos T.V. and M. K. Ravindra. 1978. Properties of Steel
(9).
(8).
for
Use
in
LRFD,
J. Struct.
1459-1468.
W.
1948. Die Temperaturabhangigkeit des Elastizitatsmoduls reiner Metalle, Z. Metallkd. 39(1).
1-9.
Uddin
T.
and C. G. Culver. 1975. Effects of Elevated Temperature on Structural Members.
ASCE.
14
101
(7).
J. Struct.
Div.
1531-1549.
NISTNCSTAR
1-3D,
WTC Investigation
Elastic Properties
Table 2-1. Specimen data for Young's modulus (E) determination
Tower Column
Specified Yield
Line and Floor
Point
Range
F,(ksi)
Specimen
WTC
N9-C2B-FR-1
1
Location
.
in
Column
55
Right flange
50
Unknown
55
Left flange
Column 154
Floors 101-104
Unknown
C68-C3T1-EP-2
butt
from
perimeter column
plate
WTC
S9-C3T-FL-1
1
Column
133
Floors 97-100
T
1
1
I
I
I
I
I
I
I
E(T)= Gq + e^T + GjT^ + e^T^
= +206
200
GPa
= -4.326e-02
e, = -3.502e-05
GPaAC
GPa^C^
-6.592e-08 GPa/'C^
Q-
150
o
NIST
1
2 Tall
UJ
3 Stanzak
4 Brockenbrough
5 ASME [C]<0.3
6 ASME [C]>0.3
7 Clark
100
9^.
%
%
8 De-er
9 Koster austenite
1 0 Koster ferrite
400
200
600
800
1000
Sources: Clark 1953; Cooke 1988; Koster 1948; Dever 1972;
Uddin 1975:
ASME
2004.
Figure 2-1. Young's modulus as a function of temperature.
NISTNCSTAR
1-3D,
WTC
Investigation
15
Chapter 2
200
180
re
Q.
CD
m
160
140
120
200
0
400
800
600
T,°C
10:33:13
-
Friday
November
05,
2004
Figure 2-2. Young's modulus, E(T), and shear modulus, G(T) for 0°C < T< 725°C. Young's
modulus was measured on the WTC steels summarized in Table 2-1. Solid line is
Eq. 2-2. Shear modulus, G, calculated from E and v via Eq. 2-4.
0.320
0.315
0.310
0.305
0.300
0.295
0.290
0.285
200
0
400
T,
10:33:16
-
Friday
November
05,
600
800
°C
2004
Source: Dever 1972; Clark 1953; Cooke 1988.
Figure 2-3. Poisson's ratio (v) as a function of temperature. The solid line
4th order polynomial (Eq. 2-3) for 0 °C < r< 725 °C.
16
NISTNCSTAR
1-3D,
is
the
fit
of a
WTC Investigation
Elastic Properties
Figure 2-4. Fractional error in representing the Young's modulus data for the three
specimens of perimeter column steel (Table 2-1) using Eq. 2-2.
NISTNCSTAR
1-3D.
WTC Investigation
17
Chapter 2
This page intentionally
18
left
blank.
NISTNCSTAR
1-3D,
WTC Investigation
Chapter 3
Room-Temperature Tensile Properties
introduction
3.1
The National
Institute
of Standards and Technology (NIST) had three goals in characterizing the tensile
properties of the steels, bolts and welds.
The
first
was
to
compare the measured properties of the
steels
with the specifications they were purchased under. The second was to determine whether their properties
were consistent with the properties and quality of other
(WTC)
The
construction era.
third
was
structural steel
from the World Trade Center
to determine the constitutive behavior for input into the finite
element models of the Investigation, which modeled the response of the building components to the
airplane impact and to the high temperatures produced
To
satisfy the first
and
by
the subsequent fires.
and objectives, NIST tested several hundred specimens taken from columns,
truss seats that represent all the
major strength levels and most of the grades of steel in the
trusses,
fire
and
impact zones of the two towers.
To
satisfy the third objective,
for each of the
many
NIST produced
values yield strengths and representative stress-strain curves
Team
grades of steel in the towers and supplied these to other Investigation
members. The representative
drew on experimental data
tensile stress-strain curves
as well as surviving
were corrected
for
dynamic
The curves
The Voce
effects.
WTC mill test reports and literature data.
work-hardening law (Voce 1948) described behavior of the representative curves. The parameters for the
Voce law
originated from the tensile behavior of recovered
3.2
TEST procedures
3.2.1
Steel
WTC steels.
Standard tensile tests were generally conducted according to
ASTM International (ASTM) E
8-01,
"Standard Test Methods for Tension Testing of Metallic Materials" using a closed-loop, servo-hydrauhc
or a screw-driven, electro-mechanical test machine. In
at a
some
tests the
cases, yield strength determination
was made
extension rate was increased to produce a plastic strain rate of 0.05 min"' for tensile
strength determination. In other cases, the entire test
tests
that
most
running crosshead rate that produced post-yield strain rates between 0.001 min"' and 0.005 min"'. In
were conducted
conform
Much
to
at a slightly
ASTM A
of the recovered
clearing of the
WTC
site.
at a
constant extension rate. Several
higher extension rate 0.0325 mm/s, which produced plastic strain rates
370-67, the
steel
was conducted
test
was damaged
method used by
in the collapse
steel mills to qualify steel.
of the towers or subsequent handling during
Specimens were selected from the
least
deformed regions. Even
so,
some of the
resulting stress-strain curves of the low-strength steels do not exhibit a yield point. In such cases, yield
strength
was determined by
the 0.2 percent offset
The geometrical and thickness
1-3D,
WTC
as specified in
ASTM E
8-01.
differences in the various components did not allow for a single, standard
specimen geometry for the tensile
NISTNCSTAR
method
testing.
Investigation
Whenever possible, specimens were made using
the full
19
Chapter 3
thickness of the material, according to
a mixture of the standard
were generally the
round
tensile
flat
ASTM E
For plates
8.
ASTM standard 0.500
in.
the specimens
in. thick,
diameter round geometry. Truss rods were machined into
specimens while truss angles were machined into subsize
and core columns were machined
specimens were
less than 0.75 in. thick, the
and round geometries. For plates greater than 0.75
into rounds. Figures 3-1 to
flat tensile
specimens. Truss seats
3-11 show the various
tensile
specimen
configurations.
Specimen
was measured using
strain
class
B2 extensometers
that
conformed
to
ASTM E
83, "Standard
Practice for Verification and Classification of Extensometer System." In most cases, the operator
removed
the extensometer prior to
marks on the specimen prior
specimen
to testing
failure.
Total elongation
and again following
fracture.
was
calculated by measuring gauge
Reductions of area for the round
specimens were measured by comparing the original cross-sectional area
to the area
of the
minimum
cross section at the fracture location. Reductions of area for the rectangular and square cross section
specimens were calculated based on the parabolic shape of the fracture surface, according
ASTM E
8-01.
The notch
to note
42 of
V
sensitivity
specimens. Fig. 3-1
diameters of 0.25
of two types of core column
1,
in.,
steels
was evaluated using notched round
tensile
taken from the quarter depth of plates. All three specimens have identical minor
but each has a different notch radius. Specimens were instrumented using an axial
extensometer mounted on the major diameter and a diametral extensometer mounted on the notch root.
Specimens were loaded
at
0.00033
in./s,
which produced an
initial, elastic
stressing rate of approximately
0.6 ksi/s.
3.2.2
Bolts
Tensile tests for determining the room-temperature load-displacement relation on the
generally followed
A
325 bolts
ASTM Recommended Practice F 606, "Test Methods for Determining the Mechanical
Properties of Externally and hitemally Threaded Fasteners, Washers, and Rivets." Although
columns were recovered, very few contained undamaged fasteners suitable
only one
A 490
and seven
A
325 bolts were suitable for
testing.
Of all
many
for mechanical testing. In
all,
the perimeter columns recovered,
A 325 bolts from the endplates. These seven bolts
A 490 bolt tested came from a "bow-tie" section from
floor 7. A cormnercial metallurgical testing firm tested three A 325 bolts and the A 490 bolt. To provide
load-displacement data and assess the magnitude of strain rate effects, NIST tested four A 325 bolts: two
only one, the unidentified C-68, had any associated
were the only
A
325 bolts characterized. The single
bolts at each of two displacement rates.
displacement) and as
fast as the
rather than the four required
sections
showed
by
rates
chosen were very slow (0.0007
(2.0 in/s).
For these
tests,
in/s
crosshead
two threads were exposed
ASTM F 606. Forensic observations of recovered perimeter column
that failed bolts
unambiguous load-displacement
20
The
machine would go
had from zero
data, the
NIST
to four threads exposed. In addition, to obtain
tests
did not employ the
wedge
required by
NISTNCSTAR
1-3D,
A
WTC
325.
Investigation
Room-Temperature Tensile Properties
8.00
Specimen thickness
Uniforr^i
Dimensions are
in
Inches
Test Section
Flat 2"
is
original section
thickness
Figure 3-1. Flat tensile specimen typically used for standard room-temperature quasistatic tensile tests.
5.
CO
2.500
o
ol
Specimen thickness
Uniform Test Section
Dimensions are
in
Inches
is
original section
thickness
Flat 1"
Figure 3-2. Flat tensile specimen typically used for elevated-temperature tensile tests.
NISTNCSTAR
1-3D.
WTC
Investigation
21
Chapter 3
4.10 -
4.25
Specimen
Uniform Test Section
Dimensions are
in
Inches
Flat-C
Figure 3-3. Flat tensile specimen used for
thickness is
original section
thickness
1 "
room and elevated-temperature
tensile tests.
3.94
Specimen
thickness
0.118
in.
is
(3
mm)
Dimensions are
in
Inches
Flat-C2 1"
Figure 3-4. Flat tensile test specimen used for
22
some creep
NISTNCSTAR
tests.
1-3D,
WTC
Investigation
Room-Temperature Tensile Properties
Dimensions are
in
inches
Figure 3-5. Flat tensile test specimen used for
some creep and
elevated-
temperature tests.
Norm.
3.0— Inches Overall Length
Uriform Test Section
Dimensions are
Figure 3-6.
NIST NCSTAR
1-3D.
Round
tensile
WTC Investigation
in
Inches
Rd
1"
specimen used for room-temperature and elevated
temperature tensile tests.
23
Chapter 3
Uniform Test Section
Dimensions are
Figure 3-7.
Round
in
Rd
Inches
tensile
specimen used
2"
for
room-temperature tensile testing.
4.3
oo
00
d
Specimen thickness
is
Uniforrri
Dimensions are
in
Inches
0.172
in.
Test Section
Fiat-W
1
Figure 3-8. Flat tensile specimen typically used for tensile testing of all-weld metal
specimens.
24
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
Uniform Test Section
1.125
5.0
Dimensions are
in
inches
Heat affected zone weld specimen HAZ-1
Figure 3-9. Heat affected zone tensile test specimen with flange/web weld intact. The
flange is the specimen portion that is in tension.
Uniform Test Section
1.125
Weld
Removed
and Surfaced
5.0
Dimensions are
in
inches
Heat affected zoHB wsld Specimen
HAZ-2
Figure 3-10. Heat affected zone tensile specimen with weld and web machined flush to
the flange surface. The flange is the specimen portion that Is in tension.
NISTNCSTAR
1-3D,
WTC
Investigation
25
Chapter 3
3/4-10 "breads
Typ.
NRB
Dimensions are
in
both ends
NRB
- 1
-
NRB
2
-
3
Inches
Figure 3-11. Notched tensile specimens.
Welds
3.2.3
Tests on specimens of all-weld metal from perimeter columns followed
ASTM E
8
and used
flat
specimen type Flat-Wl", Fig. 3-8.
Transverse tensile tests on perimeter column welds also followed
K-1 and K-2, Figs.
(HAZ)
3-9 and 3-10.
In these flange specimens, the
are oriented transverse to the pull direction.
The reduced
base metal (plate), through the weld and surrounding
HAZ, and
ASTM E 8 and used specimen types
weld and the two heat affected zones
section region extends
from within the
well into the base metal on the other side
of the web. The type K-1 specimen retains a stub of the perimeter column outside web about 0.5
which extends
web
is
at a right
removed
angle from the center of the specimen. In the type K-2 specimen, the stub of the
strength of the resistance welds in the floor trusses
chord of a
truss.
A push rod gripped in the upper hydraulic
hydraulic test machine loaded the truss rod at 0.001
26
long,
entirely.
The shear
clevis, Fig.
in.
in./s
was evaluated on a segment of the lower
grip of a 220,000 Ibf closed-loop, servo-
while the angles of the lower chord rested on a
3-12.
NISTNCSTAR
1-3D,
WTC
Investigation
Room-Temperature Tensile Properties
Source: NIST.
Figure 3-12. The resistance weld shear strength test
NISTNCSTAR
1-3D,
WTC
Investigation
(a)
before loading,
(b) after failure.
27
Chapter 3
3.3
RESULTS
3.3.1
Steel
Figure 3-13 displays some example stress-strain curves for perimeter column, core column, and truss
steels.
Appendix
A of this report
and reductions of area
summarizes the measured yield and
specimens
for all the
tested.
It
tensile strengths, total elongations,
also contains stress-strain curves for most
specimens. Sections 3.4.1 and 3.5.1 analyze these results in terms of engineering specifications.
Figure 3-14 shows stress-strain curves for two sets notched bar tension tests on core column
steels.
Specimen C71-1-F2 is from the flange of a F, = 36 ksi 12WF190 shape from WTC 1 column line 904
between floors 77-80, oriented longitudinally. The un-notched specimen, labeled R = <x), has an
unexpectedly high yield strength and lacks the expected yield point behavior, probably from damage
incurred in the collapse and subsequent recovery. The flange was 1.736 in. thick. Specimen B6152-1-F1
is from a flange plate of a F, = 36 ksi type 380 box column from WTC 1 column line 803 between
floors 15-18.
Stresses
The
original plate
was 2.06
were calculated on the 0.25
in.
in. thick.
diameter reduced cross section. The axial strain was calculated
using the notch height, 2R, as the gauge length, see Fig. 3-1
the 0.25
in.
reduced cross section diameter, which
is
1.
same
the
The diametral
for the three
strain
was calculated using
specimen geometries.
Figure 3-14 plots the diametral strains in the negative direction and includes the results of tests on
unnotched specimens for comparison. Because the operator sometimes removed the extensometers before
failure, the strains
may
was no evidence of a
radius
shows
not represent failure strain. All fractures occurred
transition to cleavage fracture.
that neither steel
by
ductile hole growth; there
tensile strength with decreasing notch
notch-sensitive under the test conditions.
is
Bolts
3.3.2
Table 3-1 summarizes the results of tests on
four
The increasing
bolts.
Figure 3-15 plots the load-displacement data for the
WTC A 325 bolts tested. All seven A 325 bolts failed in the threads, as expected.
Welds
3.3.3
Perimeter Columns
The
yield strengths of two specimens of all
specimen N-9
Fy = 84.8
ksi,
Transverse
(WTC
1,
column 154,
floors
weld metal produced by submerged
101-104 specified F,
Table 3-2. Their tensile strengths are
tests
on weld specimens were conducted on
142, floors 97-100, specified F,
type, Tables 3-3
TS^
= 60
ksi).
The
103 ksi and TS
fillet
arc welding
from
55 ksi) are F, = 85.1 ksi and
=
103.3 ksi.
welds from specimen N-8
(WTC
1,
column
yield and tensile strengths are independent of the specimen
and 3-4, so the presence of the stub of the web did not create
Because none of these transverse specimens broke
strength, but rather represent the strength of the
in the
a stress concentration.
weld, the strengths do not represent the weld
weakest region across the entire joint, which encompasses
base metal to weld to base metal.
28
NISTNCSTAR
1-3D,
WTC Investigation
r
r
'
r
Room-Temperature Tensile Properties
120
T
— —— —
I
I
—— — —— —
1
I
I
I
I
I
I
100 -
80 -
1:
C132-TA-7-W04a F =50
2:
B6152-2-F1-2-L1 (box) F =36
C30-W1-1-05 (web) F =36 ksi
HH-FL-2-L1 F,=42 ksi
3:
20 -
4:
ksi
ksi
Core Columns
and trusses
0
_i
0.00
0.10
0.05
0.15
0.20
1
I
I
0.25
I
I
0.30
Engineering strain
120
—— —
I
-I
I
100
(0
80
in
a>
i_
^—
CO
60
_c
(D
O
O)
40
c
1:
LU
2:
3:
20
4:
5:
6:
S9-C1T-S1-L1 F =36 ksi
M26-C1B1-RF-L1 F^=50 ksi
N6-C1B1-LF-L2 F,=60 ksi
C46-C2B-FL-1-L1 F,=70 ksi
C25-C1B-FR-2-L1 F =80 ksi
ClO-ClM-Fi.-1-h F=iOO kSi
Perimeter Columns
_L
0.00
-I
0.10
0.05
I
L.
0.15
0.20
0.25
0.30
Engineering strain
09:55:03
-
Wednesday September
29.
2004
Figure 3-13. Examples of stress-strain curves for perimeter column, core column, and
truss steels. In most cases, the strains do not represent failure.
NISTNCSTAR
1-3D.
WTC
Investigation
29
Chapter 3
120
-|
1
1
1
1
1
r
n
I
I
r
I
R=0.031
"^=0.062
R=0.125
in.
in.
in.
Strain
20
Core box column B6152-1-FL-1-2
WTC
1
column 803 15-18
Fy
_1
-0.2
0.0
I
l_l
I
\
I
I
0.4
0.2
= 36
ksi
L
I
I
0.8
0.6
Engineering strain
120
— — — —— — —
^
1
\
I
I
\
I
I
\
R=0.031
I
I
in.
/
100
R=0-063
CO
w
80
R=0,125
in.
in.
0)
00
O)
60
Axial
0)
Strain
<X)
40
c
LU
20
Core wide-flange C71-1-F1
WTC
1
Column 904 77-80
Fy
I
0.0
-0.2
0.2
I
0.4
I
I
I
= 36
I
I
ksi
L
0.6
0.8
Engineering strain
10:25:34
-
Wednesday September
29,
2004
Figure 3-14. Stress-strain curves for notched round bar tests using the specimens
in Fig. 3-11.
30
NISTNCSTAR
1-3D,
WTC Investigation
—
I
Room-Temperature Tensile Properties
Table 3-1. Results of tensile tests on bolts.
on
Tests
A 325 Bolts
1
ensue
Strength
Test Type
A<sTM
F
r
i
i>-i
Bolt Size
n"*
uvju-u—
f,C\fi
/
o in.x4.zj
Notes
(Ibf)'
failed in threads
in.
A<sTM
F fiOf^
D"'
uuO-u_
/\^
I Xvi r
/
o in.x4.^ J in.
A^sTM
F uuu
f,Of^-()'>
1 iM r
u—
1
JS
in.x4._j
AO nnn
failed in threads
idiieu in inreaus
in.
Tensile
esi
1
Rate
Strength
(in./s)
(Ibf)
load-elongation
7 8 in.x4.5
m.
0.00065
67.900
failed in threads
load-elongation
7/8 in.x4.0
in.
0.00065
68,300
failed in threads
load-elongation
7/8 in.x4.0
in.
2.0
68.800
failed in threads
load-elongation
7 8 in.x4.0
in.
2.0
68.200
failed in threads
Test
on an
A 490 Bolt
TS
Test
Bolt Size
A 370 tensile'
a.
For
A
325 bolts
for 7/8
b.
For
A
in.
490
length );
c.
1
7"Sm,n
in.x5
153.500
in.
= 120,000
psi.
For 7 8
psi;
minimum
in.
Notes
(psi)"
(psi)
163,400
minimum
bolts
% (2
ROA = 59.5 %
gage length)
= 55,450
on^ =
£1, =
tensile strength
16
in.
Ibf based
0.462
in."
diameter bolts.
bolts 7"Sm,n
minimum
A 490 bolt was
= 150.000
ROA = 40
reduction in area
tested as a
F,
= 130,000
minimum EL, =
psi;
14 percent (based on 2
in.
gage
percent.
round tension specimen machined from the
8x10'
1
1
1
1
I
1
1
bolt.
1
1
1
1
1
1
1
1
1
1
1
1
1
6
-
r
\
\
\
= 0.00066
= 2.0 in./s
average
rate
2
- -
-
in./s
rate
Elastic loading 2.5x10**
0
I
0.00
I
1
I
1
I
1
I
1
I
1
0.05
I
1
1
I
1
I
1
I
1
I
0.10
1
I
1
I
1
I
1
I
1
I
1
\
1
I
0.15
displacement,
1
\
1
i
0.20
1
I
1
I
1
I
Ibf/in.
— — — — ——
1
I
1
I
0.25
1
I
1
\
1
I
1
I
0.30
In.
Figure 3-1 5. Load-displacement curves from four tensile tests of
A 325
bolts,
and the
average curve.
NISTNCSTAR
1-3D.
WTC
Investigation
31
Chapter 3
Table 3-2. Room-temperature weld properties as measured.
TS
Weld Location and Type
Perimeter column
N9
from
WTC
base metal Fy
base
WTC
=55
103.0
All weld metal
ksi test
weld
fillet
test
1
84.8
All weld metal
103.3
NIST
column 154 101-104
1
metal Fy =55
test
ksi test 2
Perimeter column higher strength steels
Truss resistance welds
in.
85.1
Comments
NIST
Perimeter column lower strength steels
5/8
(ksi)
column 154 101-104
1
Perimeter column
fromN9
weld
fillet
(ksi)
92.3
105.0
118.0
From
46.4
NIST
test
Total
Load 104,000
(in shear)
Core column weld
in shear
Test
From PC&F Procedures
81.0
Procedures
From Stanray
82.9
1
PC&F
Ibf
Pacific
Procedures
5/8
in.
Core column weld
in shear Test 2
From Stanray
84.9
Pacific
Procedures
0.5
in.
submerged
arc
weld
for core
column
From Stanray
129.6
in
Pacific
Procedures
shear
Table 3-3. Results of transverse tensile tests on welds from specimen N-8
column 142, floors 97-100 specil led Fy = 60 ksi).
(WTC
1,
,
Total
Specimen Type
Specimen
HAZ-1
N8-C3B1-RF-HAZ-2
(with
TS
Elongation
(ksi)
(ksi)
(%)
70.0
95.8
12
n/a
92.6
n/d
65.3
92.2
15
67.8
92.9
17
Comments
weld root
Broke
at
Broke
0.5
stub Fig. 3-9)
HAZ-1
N8-C3B1-RF-HAZ-4
(with
stub Fig. 3-9)
N8-C3B1-RF-HAZ-1R
HAZ-2
(without
in.
from weld
stub Fig. 3-10)
N8-C3B1-RF-HAZ-3R
HAZ-2
(without
stub Fig. 3-10)
Key:
n/d, not detennined.
Table 3-4.
Fillet
weld sizes for various plate thicknesses
Plate Thickness
Range
0.75 to 1.5
1
5/8 to
1
7/8
2 1/8 to 2 5/8
32
(in.)
in
Fillet
the core box columns.
Weld
Size (in.)
3/8
7/16
1/2
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
Trusses
Although much of truss Tl was seriously deformed,
relatively
undamaged lower chord
sections with resistance welds.
104.000 Ibf using the geometry of Fig. 3-12.
resistance
weld measure 0.45
the tu'o welds
on the side
shear strength
is
46.4
in."
was possible
it
and 0.67
that did not fail
On
in.".
to select several 5 in. to 6 in. long,
The load
the fracture surface, the
Assuming
the
for shear failure
two weld areas
two angles shared the load equally, and
have the same area as those on the side
that did fail, the
COMPARISON WITH ENGINEERING SPECIFICATIONS
3.4.1
Steel
There are a number of issues
that
weld
recovered steel that complicate answering the question, "Are the
in testing
NIST-measured properties of the
steel consistent
groups: differences between the
steel properties,
the
ksi.
3.4
main
was
make up
that
and confounding
NIST
effects
test
with the appropriate specification?" They
procedures and
fall into
three
of
those in the mill test, natural variability
from the post-fabrication handling of the
steel.
This section
first
discusses these complicating factors, and then summarizes the mechanical property data for each
representative steel specimen in relation to the requirements of the original specifications.
Test Method Differences
One
issue that complicates
comparing the NIST-measured properties of the
the original specification arises
mill test report
For structural
of the
NIST
and the
steel,
tests
test
ASTM A 6
on recovered
r, lie
that the
were conducted
strain rate in the
strain rate
by assuming
MPa/s <
uniform cross section
that all the actuator
used
in the
to 5 ksi per
E
8 tests at
point measured in the
report
from
NIST.
decade of strain
NIST
accord with
in
to
r
<
no
Another problem
that
A
in establishing the
ASTM E
ASTM A
which
8,
is
a
be 0.00104
is
s"'.
A 370 prescribes
370.
Most
more generic
E
8 prescribes
only that the
In general, the operator calculates this
elastic strain in the
not infinitely stiff
specimen, which
It is
is
a reasonable
A 370 rate used for the mill test report is between 5 and 10 times the rate
A rule of thumb is that the measured yield strength increases between 2 ksi
rate.
(Rao 1966. Johnson 1967, Barsom, 1987)
Investigation could be
rate sensitivity, so
is
11.5 MPa/s.
motion produces
up
to 3 ksi less than the
rate effects alone. Unfortunately, the "rule
exhibit almost
methodology used
Investigation team used.
frequently a poor assumption, because the test machine
assumption, however, that the
test
requirements of
differences between the two are in the testing rate.
in the range 1.15
maximum
NIST
specifies that mill tests are conducted in accord with
steel
The major
standard for tensile testing.
that the testing rate,
from differences between the
methodology
steel to the
it is
of thumb value"
is
On
the average, then, the yield
one measured
in the mill test
approximate: some steels
impossible to predict the magnitude of the effect.
370 specifies three alternative methods for determining yield point and does
not require the mill to report which method
it
used. In any case, very few mill test reports have survived,
and none exist for the particular plates or shapes that are in the NIST inventory. The three methods are
•
Drop of the beam method
•
Position of the knee or
NISTNCSTAR
1-3D,
WTC
(for quasi-load controlled systems)
first
Investigation
maximum load
in the
curve
33
Chapter 3
Total extension under load
•
In any given test, these three
yield point exists,
will generally
all
yield slightly different values for the yield point. If a definite
be the highest of the possible values.
NIST
unaware of any studies
is
document whether one of the other methods consistently yields higher values, however.
that
A
it
methods
370 requires
and shapes with
/
<
1.5 in.
must be tested
full
thickness with an 18
in.
long,
wide specimen. Because of test machine limitations and lack of suitable, undamaged material,
1.5 in.
NIST
that plates
frequently tested smaller specimens than
0.5 in.
wide
0.5
diameter round specimens. Most
in.
on the 18
in.
full
thickness
flat
ASTM A 370 calls for. NIST generally used
specimens for plates with /<0.5
long specimen using an 8
in.
For most other
plates,
8 in. long,
NIST employed
ASTM standards prescribe that total elongation is to be measured
in.
gauge length, instead of the 2
in.
gauge length used for the 8
specimen. For tests on samples taken from the same plate, the total elongations measured using 8
flat
specimens will be larger than those measured using
in the shorter
scaling law between the
During the
from
1
8 in. long specimens,
specimen will occupy a greater fraction of the
two gauge
total
in.
in.
long
because the necked region
gauge length. There
is
no absolute
lengths, however.
WTC era ASTM A 6 prescribed that specimens for mill tests of wide-flange shapes originate
a section
of the web. In some cases due to material quality problems,
NIST
harvested specimens
from the flanges of wide-flange shapes. In general, because they are thicker than the webs, and therefore
cool
more slowly,
the flanges will exhibit a yield point that can be several ksi lower than tests
from the
webs. In the 1974 AISI report on property variability, the average for the difference between product
flange tests and official mill test reports
4 ksi for higher strength
is
nearly zero for low strength (F,
< 40
ksi) steels, but is
about
steels.
Natural Variability
ASTM steel standards require very little tension testing for mill acceptance. For example A
required only two tension tests from each heat, unless the heat
was necessary. The standard does not
clarify
yield different values. In the early 1970s the
how
was
less than
the yield point shall be reported
American Iron and
36-66
50 tons, where only one
when
the
two
test
tests
Steel Institute (AISI) conducted an
extensive study on the variability of mechanical properties and chemistry of structural
steel.
That report
has a great deal of information on product tension tests as they relate to the mill test report. The critical
aspect
is
that the mill test report is a single tension test,
a check that the quality of the steel
is
the AISI report (1974) and points out that a mill
product supplied has
YP > 36
conducted on one piece of a large
lot
of steel.
It is
The contemporary version of A 6 specifically calls out
test report of YP > 36 ksi is not a guarantee that all the
acceptable.
ksi:
"These testing procedures are not intended to define the upper or lower
limits of tensile properties at all possible test locations within a heat of
steel. It is well known and documented that tensile properties will vary
with a heat or individual piece of steel. It is, therefore, incumbent on
designers and engineers to use sound engineering judgement when using
tension test results shown on mill reports. The testing procedures of
Specification A6/A6M have been found
normal structural design criteria."
34
to provide material adequate for
NISTNCSTAR
1-3D,
WTC
Investigation
Room-Temperature Tensile Properties
The AISI property
variability study (1974) has a substantial
amount of data on product
flange and plate steels, where the product tests are taken from the
same
specimen. The variability within the entire heat, which encompasses
be larger than
The chapter on wide-flange
this.
YP < 40
ksi. In
mill test report.
other words,
not
uncommon
Barsom and Reisdorf (1988) reached
A
heavyweight rolled
Effects Arising
many
plates or shapes, will certainly
same shape
is
about 2.3 ksi for shapes
for product tests to be significantly lower than the
similar conclusions
from tension
tests
on
36 wide-flange shapes.
From
All steels relevant to the
on yield
it is
on both wide-
sections indicates that the standard deviation of the
difference between the mill test report and a product test from the
with
tests
plate or shape as the mill test
Prior Deformation
WTC Investigation except the F,. =
point, YP, rather than
on yield strength
100 ksi steels were qualified
F, or YS. Figure
3-16 defines yield
at the mill
based
point, yield strength,
and yield point elongation graphically. Alpsten (1972, Fig. 5) reported a typical difference between the
yield point (the maximum stress reached before yielding) and the yield strength (for instance defined by
the 0.2 percent offset
method) as about 3
ksi.
Individual values were as large as 10 ksi. Testing at
has borne out this observation. Plain carbon steels, such as
NIST
A 36, will almost always exhibit a yield point.
A 441) may or may not exhibit a yield point, but there is no a priori method
Microalloyed steels (A 572,
to predict the existence or magnitude.
Q
I
I
I
I
I
I
I
I
I
I
I
0.00
0.01
I
I
I
I
I
0.03
0.02
I
I
I
I
1
0.04
1
1
1
1
1
0
0.05
Strain
Figure 3-16. Schematic diagram of the various definitions of yield behavior
mechanical testing of steel.
NISTNCSTAR
1-3D.
WTC
Investigation
in
35
Chapter 3
The
yield point elongation, or Liiders' strain, in Fig.
in the
gauge length. Mobile
dislocations and pin
atmosphere of
them
interstitial impurities,
in place.
interstitials.
An
overstress
3-16
because of inhomogeneous deformation
arises
mostly carbon and nitrogen, can diffuse to the cores of
is
necessary to break the dislocation free of
its
Rather than homogeneous deformation throughout the gauge length, the early
sweep through the gauge
stages of deformation of such steels proceeds as deformation bands
section.
A
pre-strained specimen, where dislocations have already been freed from their interstitial atmospheres, will
not exhibit yield point elongation.
Any
The
prior deformation
remove the yield point
what the mill
may be
sufficient to
would have
stated.
or less than what the mill test
steel. In this case, the
would have
steel
reported, and
steels should exhibit a yield point
NIST-measured yield
may
it
and yield point elongation
if
a significant probability that the
is
than the specification called
for.
be
strength
less than the specified
undeformed.
does not exhibit yield point behavior, and the resulting yield strength
offset yield strength, there
still
\
•
36
Two
point.
the yield point, but only partially
Larger prior deformation will eliminate the yield point
value.
A
remove
elongation. In this case the NIST-reported value will almost certainly be less than
test report
elongation entirely and result in a work-hardened
may be more
The
removes the yield
prior deformation of the material for the test specimens typically
scenarios exist.
is
If a low-strength
expressed as 0.2 percent
NIST-reported yield strength will be
less
This does not imply, however, that the NIST-measured mechanical
properties of the steel are inconsistent with original specification. In general, a fraction of a percent strain
will
remove the
yield point, and a further (1-3) percent will
remove
the yield point elongation.
The
strains
induced by the collapse and subsequent handling are probably sufficient to remove the yield point
elongation.
Any
test
of a
F,.
= 36
ksi steel that
does not resuh
in the
appearance of a yield point and yield
point elongation should be regarded with caution.
The
pre-strain that
removed the yield point elongation can
further reduce the
measured yield strength
through the Bauschinger effect (Uko 1980). The measured yield strength of a specimen that has been
strained in the positive direction and subsequently tested in the negative direction will be less than
originally
measured
in the positive direction.
direction of the tensile test
is
The
yield strength of a
deformed flange may be lower
opposite to the direction of its pre-strain.
The
literature is not clear
if the
on the
possible magnitudes of the reduction for pre-strains less than the typical yield point elongation, however.
Identification
The
results
Methodology
of the NIST-conducted chemical and mechanical characterizations can be compared
requirements of the various
ASTM
steel specifications
allowed in the
steel contracts. In
some
to the
cases,
is
it
possible to determine with reasonable confidence that the properties of the steel are consistent with the
requirements of individual
ASTM or proprietary
specifications. In other cases, especially for the
perimeter columns where the historical record indicates that
chapter attempts to identify the specific mill that rolled the
many
proprietary steels were used, this
steel. Identification
of the specific mill cannot
be made with the same level of certainty as correspondence with intended specifications, because
requires combining information from several disparate sources and
on
to
historical data
and steelmaking practice.
be fabricated from
steel
meeting a given
No
ASTM
structural engineering plans specify only the
36
documents
it
making reasonable assumptions based
exist that identify
whether a given column was
standard or proprietary specification. Instead, the
minimum yield
strengths of the steel.
The
NISTNCSTAR
fabricator
1-3D,
was
free
WTC Investigation
Room-Temperature Tensile Properties
to
choose from a large
Fy = 36
ksi.
list
ASTM A 36
of acceptable
steels.
One
exception to this rule
the only structural steel specification in the
is
is
list
for
columns with specified
of approved
steels that supplies
with this yield point.
steel
Figure 3-17 illustrates the method for identifying possible steel grades graphically. The Port of New
York
Authority'
enumerated
(PONYA)
ASTM
contract with the steel fabricators allowed
fabricators could request permission
data.
them
to
use steels that met
standards, as well as specific proprietary steels, without further approval. In addition,
from
PONYA to use certain unlisted steels after providing property
The Port Authority of New York and
New Jersey (PANYNJ)
documents reviewed during the
many examples of these requests. These documents do not provide sufficient
unknown steel, however, because the standards generally only specify the
allowable ranges of a few chemical elements. Steel mills have much more specific recipes for individual
steels. Occasionally, NIST has identified literature examples of named steels that can help identify
unknown steels. The boundaries of the diagram should be considered porous, since NIST has incomplete
Investigation contain
information to identify an
information on construction era approvals of steels, as well as chemical and mechanical characterization
of proprietary
properties.
state
steels.
From
and the
with certainty that a specific mill rolled a specific plate.
one mill supplied
was
Furthennore, as discussed above, each steel exhibits a natural variability in
the chemical and mechanical analyses
likely to
a proprietary steel, the analysis discusses
have met
historical record,
Where
it
is
never possible to
the historical record indicates that
whether the specimen
in the
NIST
inventory
that specification.
Figure 3-17. Methodology for identifying recovered steels.
The
rest
of this section summarizes the representative specimens and attempts to discern whether
likely that the steel
met the
detailed above induce.
A
was
original specifications, subject to the constraints that the complicating factors
separate
steels characterized as part
it
document (NIST
NCSTAR
1-3B) summarizes the chemistries of all
of the Investigation. Another document (NIST
NCSTAR
1-3 A)
summarizes
the standard specifications for the steels used in the construction.
Perimeter Columns
Contemporaneous construction documents (White 1967; Symes 1968) indicate
Foundry (PC&F) fabricated the perimeter columns primarily with
and
Steel,
and possibly Kawasaki
NIST NCSTAR
1-3D,
WTC
Steel.
Investigation
steel
it
that Pacific
Car and
purchased from Yawata Iron
One contemporaneous document (White 1967)
indicates that
37
Chapter 3
Yawata was
least
> 36
to supply the steels with F,
ksi
and Kawasaki was
one surviving shipping manifest (Mitsui 1968) shows
that originated in Chiba, Japan, the location
document
Yawata
clearly indicates that
PC&F
of a Kawasaki Steel
also supplied
A
t
= 36
ksi.
At
of A 36 spandrel plates
One contemporaneous
mill.
construction
36 for spandrels (Symes 1969b).
Another contemporaneous document (Symes 1969) also indicates
plate 3, the inner
to supply steels with F,,
received 190
that
PC&F
web, of the perimeter columns. By weight, the plate 3
used domestic
steel for
steel represents less than
10 percent of the mass of the perimeter columns, because the spandrel plates
make up much of the
inside
wall of the perimeter column.
Except for the
to
its
own
F,.
= 36
ksi steel,
specifications,
which
which the
complete specification for these
supplied to meet
it
A 36, Yawata supplied all other grades of steel
PONYA chief engineer approved. NIST was unable to locate the
steels.
The only document
that
NIST
located (White 1967b) details
required changes to the complete specifications for Yawata steels during the negotiations over acceptance.
Unfortunately,
NIST
could not locate the original specification document that describes the tensile
strength and total elongation requirements of any of the
Yawata
impossible to compare the measured tensile strengths and
total
steels
with F, > 36
ksi.
It is
therefore
elongations of perimeter column steels to
specification documents. Generally, the measured total elongations of the intermediate-strength steels are
greater than the 24 percent
In general, the
Figure 3-18
tests
minimum
that
specified.
measured yield strengths perimeter column
plots the ratio
of the measured yield strength
steels
minimum requirements.
minimum yield strength for
exceed the
to the specified
all
using specimens oriented parallel to the original rolling direction of the plate, which was the
specified test orientation for plates in the
unity ratio
at
specified F,
panel, and thus probably
38
A 441-66
= 60
WTC construction era.
ksi originated in
came from
the
same
two inner web
The
plates
points that
lie
below the
line
of
from the same perimeter column
plate.
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
1.8
values <
Perimeter Columns
L orientation only
1 arise from
natural variation
>-1.6
LL
A Plates 1,2,4
o Plate 3
D
.0
0
F =(55 60 70 100)
1.4
Q.
ksi
values offset for
clarity-
C/D
^
1.2
CO
CO
CD
^
Q
I
1.0
bars represent
max and min
points without bars represent single
values
measurements
0.8
40
60
100
80
Specified F
ksi
Figure 3-18. Ratio of measured yield strength to specified
yield strength for all longitudinal tests of perimeter
minimum
column
steels.
Fy = 36 ksi S-9 Spandrel
The
representative F,
perimeter column S-9
hole.
The
= 36
ksi specimen,
(WTC
1
Column
S9-C1T-S1, came from the 100th floor spandrel of recovered
line 133 floors
rolling, or longitudinal, direction
97-100), which
consistent with the requirements of
A
=
A
36, Table 3-5.
The
Mn
Two
and
A
in longitudinal tests
NISTNCSTAR
is
exceed the
36.
3-5
column
Si contents are higher,
1,
and the
36 has no requirements for
Mn
is
C
also consistent with the requirements of
content
is
lower, than most of the other
or Si for plates that are 0.75
of the thirteen characterized spandrels specified as Fy = 36 ksi have
the other eleven
above the airplane impact
38.2 ksi. The tensile strength and total elongations are also
The chemistry of specimen S9-C1T-S1, Table
Fy = 36 ksi spandrels, but
just
of the spandrel plates lies in their long direction, or
perpendicular the axis of the building. The measured yield strengths
requirements of A 36; the average F,
is
in.
this chemistry; the
thick or less.
chemistry of
different, but also consistent with the requirements of A 36.
1-3D.
WTC Investigation
39
Chapter 3
Summary
Table 3-5.
of mechanical properties and chemical compositions for steels
from low-strength perimeter columns.
aI4-CiD- Mzo-C 1
FR-1
S-1
1
-
1 -
iNo-Ci
FR-1
N8-C1B-F-1
OW-1
0.16
0.17
0.16
0.18
C
1.18
1.36
1.27
1.50
1.39
Mn
(note a)
ClB-FL-2
c
0.10
0.19
0.16
Mn
1.41
1.21
P
0.01
0.020
<0.005
0.024
0.016
0.023
0.018
P
S
0.016
0.016
0.024
0.022
0.011
0.016
0.012
S
Si
0.55
0.25
0.36
0.29
0.40
0.44
0.40
Si
Ni
0.02
0.01
0.06
0.01
0.01
0.01
0.02
Ni
Cr
0.04
0.03
0.05
0.02
0.02
0.02
0.04
Cr
Mo
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
Mo
Cu
0.08
0.05
0.14
0.04
0.05
0.05
0.05
Cu
0.018
<0.005
0.024
<0.005
<0.005
0.005
0.011
Nb
<0.005
0.040
<0.005
0.047
0.042
0.064
<0.005
Nb
Ti
0.026
<0.005
0.018
<0.005
<0.005
0.007
<0.005
Ti
Zr
<0.005
<0.005
<0.005
<0.005
<0.005
<0.005
<0.005
Zr
Al
<0.005
<0.005
<0.005
0.018
0.024
0.042
0.029
Al
B
<0.0005
<0.0005
<0.0005
<0.0005
<0.0005
<0.0005
<0.0005
B
N
0.0032
0.0070
0.0060
0.0186
0.0065
0.0033
0.0050
N
36
45
46
50
55
60
60
V
F, (ksi)
1
V
F, (ksi)
specified
specified
Fy avg
38.2
56.7
61.7
55.8
69.2
61.8
Mm.)
a.
b.
3/8
1
7/8
9/16
3/8
S9-C3T-S-1 and S9-C1T-S-1 originate from different locations on the same plate
longitudinal tests of N8-C1B-F-1 is Fy = 68.7 ksi. Value reported
3 tests on N8-C3B-RF
is
1/4
the average of 4 tests on
/(in.)
N8-C1B-F-1 and
lon^, longitudinal^
Note: Siighlighted entries are particularly relevant for
Fy =
steel identificationj
45ksiASCE3
The representative
at
(long)
5/16
1/4
The average of 4
Key:
Fy avg
60.8
(note b)
(long)
F,
= 45
ksi
specimen
ASCE3-C1B-FL came from an
one end identified the yield strength of the
longitudinal test, F,
total
=
plates,
steels
plates are F,
=
5
1
of similar thickness. In that group of mill
.6 ksi,
TS = 72.8
ksi; the total
A
starnping
however. The yield strength measured in the single
56.7 ksi, far exceeds the specified F,
= 45
ksi.
elongation are consistent with values from a group of recovered
Fy = 45 ksi
unidentified location.
The
yield and tensile strengths and
Yawata
test reports, the
elongation in 8
in. is EJ,
mill test reports for
averages of 25 individual
= 28.0
percent.
The chemistry of ASCE3-C1B-FL is consistent with the Yawata "A441 -modified" chemistry
specification. Table 3-7 under the heading "Yawata 50." The important features of that specification
are
the high allowable Si content and the use of Nb as a micro-alloying agent. That specification also
expresses the
40
Cu
content as a
maximum
rather than the usual
minimum.
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
Fy
= 46
ksi
S-14 Spandrel
The representative
the north face of
= 46
Fy = 46
F,
F,
= 46
specimen S14-C3B-S1 came from the 91st floor spandrel of a panel from
ksi
WTC 2 on column line 217. The structural engineering plans specify this spandrel as
ksi.
For unknown reasons, the plans include spandrel plates specified as
ksi,
sometimes even
same perimeter column
in the
Fy
= 46
ksi steel. All surviving
Yawata
= 46
or
A 441
In
any
ksi for the thickness
measured yield strength
Fy = 55.8
ksi,
which
is
ksi steel.
Table 3-5.
It
is
SHCR and PONYA
= 45
ksi
ever mention
and F, = 50 ksi
steels only.
structural steel specifications describe a steel supplied
/
=
3/8
in.
A
For example, plates meeting
ksi for this thickness.
A 572
does not have a
F,.
242,
= 46
A 440,
ksi entry.
exceeds the specified minimum; the average value of three
far
also consistent with any of the F,
The chemistry of the specimen
Fy - 50
of this spandrel:
would be supplied with Fy = 50
case, the
tests is
ASTM
and
ksi
However, none of the surviving
and
specifications describe Fy
Furthermore, none of the contemporaneous
with Fy
panel.
PC&F
contemporaneous construction communication between
= 45
F,.
= 50
ksi specifications.
consistent with the requirements for the specification for the
Yawata
does differ from the chemistry of the other two F, = 50 ksi plates
recovered, however.
Fy = 50 ksiM-26
The representative
line 131 near the
F,
= 50
ksi
specimen M26-C1T-FR came from the north face of WTC
93rd floor, which
section of flange (plate 1) 9/16
The chemistry of the specimen
column
just
minimum:
is
it
is
so
many
=
61
test
on column
1
specimen originated
longitudinal yield strength
is
in a
more than
.7 ksi.
Yawata
specification for F,
= 50
ksi.
Like
many of
strengthened by addition of Nb. The Yawata hot-rolled steels in the
common
construction correspondence denotes as
Nb+V,
F,
consistent with the
steels,
range 50 ksi < F, < 60 ksi share a
0.15 percent on
below the impact zone. The
The average measured
in. thick.
10 ksi higher than the specified
the other perimeter
is
chemistry specification, which the contemporaneous
"A441 -modified." That
specification only puts an upper limit of
different chemistries could satisfy this specification.
Fy = 55ksiy-9
The representative
line
F,
= 55
ksi
specimen
155 near the 102nd floor. The
thick.
test
N9-C1M-FR came from
The measured average yield strength
The chemistry
is
the north face of
is
consistent with the specification:
Yawata F, = 55 ksi specification, and
as F, = 50 ksi and F,. = 60 ksi.
consistent with the
perimeter column steels specified
WTC
specimen originated from a section of flange
is
F,.
=
1
column
(plate 1) 0.25 in.
61.8 ksi.
very similar to those of
Fy = 60 ksiN-8
The representative
F,
specimen N-8, which
100. This panel
= 60
is
ksi
specimens came from several of the flange and outer web plates
from the north face of WTC
than the specified
minimum.
1-3D,
WTC
in.
in
centered on column line 142 between floors 97-
was just above the impact zone. The average yield
specimens taken from two different 5/16
NISTNCSTAR
1,
thick flanges (plate 1)
strength of seven tests of longitudinal
is F,,
=
69.2 ksi, which
is
much
greater
The average yield strength of three longitudinal specimen taken from the
Irivestigation
41
Chapter 3
web
outer
of N8-C3-B1 (column line 141, near the bottom)
(plate 2)
was shghtly lower than
the specified
The chemistries of the flange
= 60
specification for F,
is
F,
= 60
ksi vs. F,
= 60.8
One of these
ksi.
tests
ksi).
(plate 2) are consistent with the
Yawata
Table 3-7, but they are not identical. In particular, the outer web
Nb, and the
specification only limits the
= 59.2
(F,
and outer web
(plate 1)
ksi steel.
(plate 2) does not contain
minimum
V content is rather low
sum of Nb+V, both
(0.01
percent).
1
But because the Yawata
chemistries are consistent with this specification.
Fy = 65ksiN-99
The representative F, = 65 ksi specimen came from the flange (plate 1) of specimen N99-C3T1-RF,
which originated in WTC 1, column line 147 near the 102nd floor. This location is several floors above
the impact zone. The average measured yield strength of two longimdinal specimens is F, = 71.8 ksi,
which exceeds the minimum
F,
= 65
ksi in the
The chemistry of N99-C3M-FR-1, which
with the Yawata
denote as
and
it
WEL-TEN
60.
The
from
specification.
a different location in the
same
plate,
is
also consistent
65 ksi specification. Table 3-6, which contemporaneous construction documents
F,.
was only used
WEL-TEN
is
Yawata
60R. According to these documents,
at
thicknesses up to 0.5
in.
WEL-TEN 60R is
hot-rolled
this
was
the highest strength hot-rolled grade,
Thicker plates required the quenched and tempered
alloyed with
V
instead of Nb, and has added Cr.
Fy = 70 ksi C-46, S-1, and S-14
NIST
7/16
tested three different
in.
a 3/4
in.
examples of F, = 70
thick flange (plate 1)
thick flange
thick flange (plate
1
line
from the north face of WTC 2 column
1 )
= 70
The chemistries of all
WEL-TEN
perimeter-column
from west face of WTC 2 column
longitudinal yield strength
specification for F,
ksi
from the west face of WTC
is F,.
ksi
column
They
are
S1-C2M-FL,
a
433 near floor 89, C46-C2M-FL,
157 near floor 69, and S14-C3M-FR, a 1/4
line
77.7 ksi, which exceeds the
WEL-TEN
steels.
line
218 near floor
minimum
92.
in.
The global average
requirements of the Yawata
62.
three specimens are similar. Table 3-6. All are consistent with the F,
62 specification. Table 3-7, which has
= 70
ksi
V and Cr additions.
Fy = 75 ksi C-22
The
representative F,
north face of
WTC
1
= 75
ksi
column
specimen C22-C2M-FL came from a
line
157 near floor 94. This location
average longitudinal yield strength,
WEL-TEN
=
1/4 in. thick flange (plate 1)
just
from the
above the impact opening. The
83.7 ksi exceeds the requirements of the F,
= 75
ksi
62 specification.
The chemistry
is
consistent with the
the chemistry of the F,
42
F,.
is
= 70
F,.
= 75
ksi
WEL-TEN
62 specification. Table 3-6, and
is
similar to
ksi steels.
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
Summary
Table 3-6.
of mechanical properties and chemical compositions for steels
from high-strength perimeter columns.
N99-C3MS1-C2M- C46-C2M- S14-C3M- C22-C2M- C25-C1B-
MlOb-
FL-1
FR-1
FL-1
FR-1
C3IVI-FR-1
0.13
0.12
0.13
0.15
0.13
0.14
C
1.25
1.08
1.10
1.07
1.13
0.97
0.84
Mn
P
0.018
0.008
0.014
0.010
0.016
0.014
0.009
P
S
0.018
0.013
0.015
0.013
0.013
0.015
0.016
S
Si
0.38
0.41
0.40
0.39
0.39
0.33
0.23
Si
Ni
0.01
0.01
0.04
0.01
0.01
0.01
0.01
Ni
Cr
0.19
0.22
0.21
0.21
0.23
0.32
0.88
Cr
< 0.01
<0.01
0.20
0.40
Mo
Cu
FR-1
(note a)
FL-1
c
0.16
Mn
Mo
<0.01
<0.01
< 0.01
Cu
0.05
0.04
0.05
0.05
0.04
0.28
0.23
0.020
0.024
0.020
0.041
< 0.005
<0.005
V
Vj
0.053
Nb
<0.005
< 0.005
< 0.005
< 0.005
<0.005
0.024
<0.005
Nb
Ti
<0.005
< 0.005
< 0.005
< 0.005
<0.005
< 0.005
<0.005
Ti
Zr
<0.005
< 0.005
< 0.005
< 0.005
<0.005
< 0.005
<0.005
Zr
Al
0.031
0.110
0.097
0.110
0.110
0.100
0.075
Al
B
<0.0005
< 0.0005
< 0.0005
< 0.0005
<0.0005
0.0016
<0.0005
B
N
0.0120
0.0060
0.0070
0.0070
0.0020
0.0070
0.0051
N
65
70
70
70
75
80
100
F,
71.8
77.9
75.9
79.4
83.7
95.7
107.7
1/4
7/16
3/4
1/4
1/4
F, (ksi)
Fy avg(long)
?(in)
a.
j
N99-C3M-FR-1 and N99-C3T1-RF are from different
occasionally "FR" and "RF" are used interchangeably
Kev:
long, loncitudinal.
—
c~
Note:
1/4
1/4
same
locations in the
to indicate
plate.
In the
(ksi)
F, avg(long)
l(in)
specimen designation system,
specimens from the right flange.
—
1
!Hij;hli_^hted entries are particularly relevant for _steel identifi_c_ationj
Fy = 80 ksi C-25
The representative
the east face of
WTC
specimens, F, = 95.7
WEL-TEN
= 80
F,
1
specimen C25-C1B-FR-2 came from a 1/4
column
line
207 near floor
ksi, is significantly
89.
The average
higher than the
minimum
identifies
it
as a
chrome-molybdenum
thick flange (plate 1)
from
longitudinal yield strength for two
requirement of the Yawata
steel, is also consistent
70 specification, Table 3-7 The chemistries of three other
are similar.
in.
70 specification.
The chemistry, which
TEN
ksi
A
thicker
F,.
= 80
ksi section,
C46-C1B with / =
amounts of V instead of the small amounts of Nb of the other
WEL-TEN 70 has no limits on Ni, Nb,
with the WEL-TEN 70 specification.
NISTNCSTAR
1-3D,
WTC
Investigation
F,.
5/8
F,.
in.
= 80
ksi plates
with the
from
WEL-
different
columns
contains 0.43 percent Ni and small
= 80
ksi steels.
The
or V, however, so the chemistry of C46-C1B
specification for
is
also consistent
43
Chapter 3
Fy = 85 ksi
Fy = 100 ksi M-lOb, C-10, B-1043, and C-14
to
PC&F requested
Contemporaneous construction documents (White 1968) indicate
that
permission to use F, = 100 ksi
steels
85
ksi.
steel
wherever the plans called for
Forensic investigation of the recovered steel confirmed this finding. Ten of the fourteen columns
of recovered
steel that
were specified
have
to
F,.
> 85
tested four different specimens of nominally Fy
Specimen M-10b-C3Bl-RF came from
column
line
205 near floor
83.
markings on the plate indicate
flange from the west face of
= 100
= 100
ksi
had legible yield strength
ksi.
ksi perimeter
1/4 in. thick flange (plate 1)
column
steel.
from the north face of WTC 2
Although the stmctural engineering drawings specify
it
as F,.
= 85
ksi,
= 100
ksi. Specimen ClO-ClM-FL came from a 1/4 in. thick
is F,
column line 452 near floor 86. Although specified as F, = 90 ksi,
it is F, = 100 ksi. Specimen C14-C1-RF came from a 1/4 in. thick
that
it
WTC
markings on the plate indicate
> 90
ksi or F,
stampings. In each case the columns were stamped as F,
NIST
and received
with yield strengths greater than
1
that
comer of WTC 2 column line 301. Specimen B1043-C2T-FL is an example of a
= 1.125 in., specified as F,. = 100 ksi from the south face of WTC 2 column line 406
flange from the northeast
thicker section,
/
near floor 43.
As
a group, the average yield strength
the yield strengths exceed the
Table 3-6 and Table 3-7. The
is
minimum
total
= 109.2
F,
minimum of the group
ksi; the
requirement of F,
= 100
ksi for
elongations to failure for specimen
the other specimens: 16 percent and 17 percent
compared
to
is
F,
=
107.5 ksi. All
WEL-TEN
Yawata
C14-C1-RF
80C,
are less than those of
an average of 22 percent for the others.
= 100
Some
WEL-
contemporaneous construction correspondence (White 1967b) indicates
that the F,
TEN 80C
was
A 514-66, which mandates a
minimum
total
(WEL-TEN
plate
is
to
meet the strength and
total
elongation requirements of
elongation of 18 percent in a 2
in.
gauge length.
A
1966 Yawata
steel
ksi
advertisement
1966) states a 16 percent minimum. The total elongation of each specimen from F, = 100 ksi
larger than this value.
Inner Web (Plate
3)
Contemporaneous construction documents (Symes 1969) indicate
flanges, outer
webs and spandrels
(plates
web
from domestic
For
(plate 3)
steel.
1,
2 and 4)
that
from Japanese
this reason, this section
PC&F
generally fabricated the
steel (primarily
Yawata) and the inner
reviews the properties of steel specimens
from the inner webs separately. Chemical analyses of specimens from the inner web confirm
frequently different from the steel used in plates
is
the inner
Yawata
web
steel.
plate
is
1
and
identical to the other plates in the
2.
In
some
that the steel
cases, however, the composition of
column, indicating
that
it
was
fabricated frorh a
web plates with F,. > 75 ksi have chemistries consistent with the Yawata grades.
inner web represents less than 0 percent of the total steel used in the perimeter
All inner
In terms of mass, the
1
columns, because the spandrel plates cover part of the inside of the perimeter column, and because the
inner
web
is
often thinner than the flange plates.
Other contemporaneous construction documents provide evidence that Bethlehem
the domestically produced steel for the
steel supplied
most of
PC&F perimeter column contract. The actual fraction is unknown,
that PC&F purchased or requested approval to
and those same construction documents also show
purchase plate from nearly every domestic
Fy < 100 ksi contain Nb, which means
44
steel
that they
company. None of the recovered
do not correspond
to
USS
plate 3 steels with
Ex-Ten, which specifies a
NISTNCSTAR
1-3D,
WTC Investigation
;
Room-Temperature Tensile Properties
minimum Nb coment. Because of the
lack of information in the construction record,
unambiguously identify the source of the
used for plate
steel
it is
impossible to
3.
Summary
Table 3-7.
of mechanical properties, chemical compositions, and relevant
specifications for steels from high-strength perimeter columns.
ASTM and Yawata
WFTI
>^ t L- TP\I
I iLiy
FL-1
C2T-FL-1
c
0.17
0.15
Mn
0.89
P
S
W tLj-
1
Yawata50
62
70
80C
0.22
0.22
0.18
0.18
0.18
0.80
0.85-1.25
1.1-1.6
1.40
1.20
0.6-1.2
0.015
0.005
0.040
0.040
0.035
0.030
0.030
P
0.012
0.012
0.050
0.050
0.035
0.030
0.030
S
U._4
n
Ti
U.13
U.J J
A
^C
U.J J
A
<C
U.J
J
A
^ A 7^
U.l J—U.JJ
U.U
U.U 1
A
1A
U. oU
A
1A
U. jU
c
Mo
1
n
on
u.vu
U.oo
A
U.J J
/I
7
U.zo
U.UU
V
A
U.J
U.jU
<U.UUj
U.
J
J
IN
0.6
max
1
0.12
<-U.UUj
max
A
^ A ^
U. 1 J— U.J
U.uz min
C
Mn
1
.3
1
A
OA
U.ZU
u.zu
1
<U.UUj
/
441
A
U. 7 — 11
l .J
l
0.6
MO
max
A
A ^
U. 1 J— U.J
Lu
1
max
V
D+ V
D
In
max
Ti
<0.005
<0.005
Ti
Zr
<0.01
<0.005
Zr
Al
0.081
0.053
Al
B
<0.0005
<0.0005
N
0.0040
0.0080
100
100
108.7
110.3
F, (ksi)
F,
avg (long)
/
Key:
1/4
(in.)
0.006
max B
N
50-60
50
75-80
70
100
F, (ksi)
F, avg (long)
1/8
1
/(in.)
long, longimdinai.
N9-C1T1-IW
line
is
specimen of plate
a
155 near the 102nd floor. The original plate
longitudinal specimens
steel
3 with specified F,
has slightly too
is
F,
much
=61.3
Mn
to
ksi,
which
is
is
0.25
= 55
ksi
in. thick.
from the north face of WTC
The average
larger than the specified
F,.
column
1
yield strength of three
= 55
ksi
Chemically, the
meet the Bethlehem V-series specification (1.31 percent
vs. a
maximum
1.25 percent) even after accounting for the relaxed requirements of product testing, assuming the
increased compositional ranges allowed by
effect
on mechanical properties.
NIST
tested three
the north face of
examples of F, = 60
WTC
1,
column
specimens of N8-C2M-IW
is
F,
line
A
572-70 for Mn. This difference would have an insignificant
ksi plate 3 steel.
N8-C1B1-A-IW
is
a specimen of inner
web from
143 near floor 98. The average yield strength of two longitudinal
56.7 ksi, which
is
below the specified minimum: F, = 60
ksi.
The
average of three transverse specimens, an orientation which would not have been used for mill testing
similarly low: F,
= 56.8
ksi.
N8-C2M-E-IW came from
the adjacent
column
strengths of longitudinal and transverse specimens are similarly low: Fy
NISTNCSTAR
1-3D,
WTC
Investigation
=
142.
The average
56.7 ksi and Fy
=
is
yield
57.2 ksi
45
Chapter 3
Given
respectively.
columns are adjacent and the inner web plate thickness
that the
two inner web
is
the same,
likely
is
it
came from the same original plate. Specimen C40-C2M-IW originated
from the north face of WTC 1, column line 136 at floor 100. The average yield strength of two
longitudinal specimens of C40-C2M-IW, F, = 64.6 ksi, exceeds the specified minimum.
that these
plates
plates are very similar to each other, as well as to the compositions of many of
The chemistries of the two
the other plate 3 specimens. All are high in V, and contain Cr and Ni.
The magnitudes of the
alloy
contents are consistent with the Bethlehem V-series specification.
Specimen N99-C3M1-IW
is
an example of an inner
web
plate with specified F,
= 65
ksi
from
WTC
1
column line 147 between floors 99 and 102. The average yield strength of three longitudinal specimens is
Fy - 66.6 ksi, which exceeds the specified minimum. The chemistry of the plate is very similar to all of
has vanadium added for strength, as well as trace amounts of Ni and Cr.
the other plate 3 specimens.
It
Specimen S14-C3T-IW-1
an example of an inner web plate with specified F, = 70 ksi from
column
line
Fy = 76.0
217 between
is
floors 91-94.
The average
which exceeds the specified minimum:
ksi,
WTC 2
yield strength of two longitudinal specimens
F,
= 70
ksi.
The chemistry
is
is
similar to the rest of
the plate 3 specimens and consistent with the construction correspondence on the chemistry specifications
Bethlehem V-series
for the
The measured
yield strengths of two of the inner
relevant question
is
steels.
is,
"If the
web
plates are less than the specified
measured yield strength of a
lower than the specified minimum,
is this
plate
measured
in the course
minimum. The
of the investigation
an anomalous or an expected event?" That question
is
the
convolution of at least two other questions:
1
.
2.
What is the distribution of yield
minimum yield strength?
What
is
strengths
from mill
test reports for a
given specified
the distribution of product test values of yield strength for a given strength in a mill
test report?
The
literature data
on variation
while reasonably large,
is
Galambos (1978) provide
in structural steel properties
not structured in a
partial
answers
Product Analysis and Tensile Properties.
how
.
manner
to question
that
1,
(Alpsten 1972, Galambos 1978, AISI 1974),
makes
would
like to
One might also attempt to
specified minimum yield strengths
2.
fundamentally different from the objective answers to questions
answer the question by calculating the probability
actually have a
inner
is
easy to use. Alpsten (1972) and
and the AISI report (1974) on "Variation of
helps answer question
frequently in construction that steels with inappropriate
but this question
it
measured (product) F,
1
and
estimate
are used,
2. Ideally,
that a plate specified as F,
= 60
ksi
we
would
56.7 ksi, the average yield strength of test specimens from the
web of perimeter column N-8.
Several 1960's era studies (Alpsten 1972, Galambos 1978) summarize the distribution of mill-test-report
One shortfall of these studies in applying them to the questions at
mean value and a standard deviation or coefficient of variation (C,.).
Because each grade of steel is supplied to a specified minimum value, the distributions of mill test report
values are non-normal. Instead, they are truncated at the specified minimum and therefore skewed to the
positive values. Estimating the fraction of plates with strength near the specified minimum is not accurate.
yield strengths for given steel grades.
hand
46
is
that they usually report just a
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
A
second problem
mills.
It is
is
that
no studies
exist that
likely that individual steel mills
examine the variabihty
between
in these distributions
have different production practices
steel
that result in different
shapes to the distributions. The reported values for the distributions usually represent the output of a
single mill.
The AISI
but
it
report on product variability (1974) appears at
codes
data in a
its
manner
glance to be the answer to these questions,
first
removes some of the
that
utility.
The
probabilities that
expresses for
it
product tests are given as the expected difference between the measured value and the mill
value. In other words,
for the
WTC steels.
knowledge of the
In that sense,
Given the uncertainties
in the
it
original mill test report value
answers question
AISI report (1974),
it
product
WTC
1
test report
is
impossible
is still
absent.
not prudent to attempt a detailed statistical
whether
calculation. Instead, the calculation should simply establish
than the specified
tension test result that is less
is
but a good answer to question
2,
is
necessary, which
minimum
it is
reasonable to have found a
in the suite
of tests that characterize the
steels.
To proceed with
the estimation, define the following variables:
Variable
Value
Definition
Fy
60 ksi
the specified
Fy
56.7 ksi
the
kp
-
^^
^
c
r..
yield strength for the plate in question,
NIST-measured (product) yield
1
F^"
.092
,
to the specified
strength
minimum
yield strength, F,
for high-strength
the coefficient of variation of the historical average of mill test reports
.
of a given specified
minimum
yield strength
Galambos (1978) estimated
and C, from Alpsten's (1973) data on Swedish
The AISI
( 1
report on variability
deviation from the official value
calculate the deviation, Ao,
974) expresses
(i.e.
of study SU/18
in the
for results
its statistics
F,.
^ 58
of product
the value stated on the mill test report).
from the mill
To
ksi plates.
tests in
terms of
estimate this value,
test report as
Act = k/^. - F/ =
1 1
,
steel plates
0.054
From Table
N-8
the historical average of the ratio of yield strength in mill test reports,
„mtr
r
minimum
1
.092 x 60 ksi
56.7 ksi
-
= -8.82
AISI report (1974), the minimum
ksi
fraction,
/
of tests higher than
this
difference from the official test is/ = 0.87 for steels with F,>50 ksi. Therefore 13 percent of the time,
product tests of plates would have produced values such as those for plates
N8-C1B1-IW. Given
specified
minimum
NISTNCSTAR
1-3D,
that they
were physically located next
yield strength,
WTC
it is
Investigation
to
likely that these inner
N8-C2M-IW
and
each other, and are the same thickness and
web
plates originated
from a
common plate.
47
Chapter 3
In comparison,
plates
was
from
NIST
more than 20
tested
N8-C2M-IW and N8-C1B1-IW,
less than the specified
yield strengths of inner
calculation.
shows
It
is
that
168 mill
test reports.
reports for steels
Of these,
made
possible that a product tension test of
made
to
web
minimum.
less than the specified
WTC
steel
could be less than the
considered anomalous. Comprehensive data on the range of possible
The data from study SU/18 of the AISI
report (1974)
That report does not contain sufficient infonnation to decide
steels
only the inner
approximate and should be regarded only as an order-of-magnitude
it is
minimum without being
came from
plates.
not unreasonable or anomalous that the measured
it is
from perimeter column N-8 are
plates
values of kp and C, do not exist.
reports
column
or about 5 percent of the total, had a product tension test that
minimum. Therefore,
web
The previous calculation
specified
distinct perimeter
to high-strength specifications, or if they
low strength
if
is
based on only
those 168 mill test
were simply very high mill
test
specifications.
Core Columns
NIST tested representative samples of the
Fy = 36
ksi
complete
and F,
test
some of the specimens
from damage
to the
column
common
core columns: box and wide-flange in
results.
Appendix
A
is
less than the specified
minimum,
of this report tabulates the
Mechanically, the
tests.
but these differences probably
removed the yield point or test differences between the mill test
3-19
Figure
plots the ratio of measured yield strength or yield point to the
that
NIST protocol.
minimum for longitudinal
report and the
specified
ksi.
four most
Table 3-8 summarizes these
data and plots the stress-strain curves for the individual tension
yield stress of
arise
= 42
tests.
Table 3-8 and Table 3-9 also summarize the chemistry data on specimens of identified core column
With one exception, the
steels are
unremarkable. The chemistry of specimens from
= 36
(shapes and plates) specified as F,
strength shapes are consistent with
contract, both
HH-Fl-1 and C-26
(a
A
572.
with every
or
A
572. Although
beam from
Bethlehem V-series requirements. That
meet
ksi are consistent with
A 441
specification listed in the
undeformed regions
slightly bent.
to harvest test
steel.
columns
ASTM A 36. The chemistries of the highA 572 is not listed in the PONYA steel
PONYA steel contract and could be sold to
yield strength of specimen
C-88c
is
inconsistent
PONYA contract, but that contract lists proprietary steels for
which NIST does not have typical compositions
Most of the core columns recovered were
the
the 107th floor) could have been supplied to the
steel is listed in the
The combination of chemistry and measured
ASTM
all
for comparison.
significantly deformed,
which made
specimens from. Even the relatively
Appendix B summarizes the extent of the deformation
it
difficult to select
straight sections
in the regions
were often
from which specimens
were taken for colunms C-30, C-65, C-70, and C-155. From the geometry of the deformed colurrm,
possible to estimate the magnitude of the changes to the stress-strain curves for the
web and
it
was
flange
specimen. Generally, the disappearance of the yield point and the elevated strengths of the flange
specimens are consistent with the magnitude of the deformation measured from the shape of the core
columns.
This section describes each column and the individual specimens harvested from
48
it.
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
Fy = 36 ksi Wide-Flange C-30
= 36
14WF237 shape from
WTC
Specimen C-30
is
This location
on the opposite side of the building and well above the impact zone. All
is
a F,
specimens, three from the
ksi
web and
three
from the
elongation behavior. All six tests produced
2
column
line 1008,
between
floors 104-106.
six tensile
flange, exhibit both yield point and yield point
YP > 36
ksi.
The segment of the column from which
specimens originated had no measurable deformation, see Appendix B, which
is
the test
consistent with the
appearance of the yield point and the similarity of the stress-strain curves from the webs and flanges.
Figure 3-20 plots the stress-strain curves for the longitudinal tests of C-30, along with the corresponding
curves for the three other wide-flange core columns specified as
F,.
= 36
ksi for
conducted. The chemistry, tensile strength, and total elongation of C-30 meet
Table 3-8.
Summary
C30-F1-1
Shapes
C
tests
were
36.
and chemical compositions for steels
from core column wide-flange shapes.
of mechanical properties
C71-FL-2
C65-F1-1
WTC 2
which web
ASTM A
WTC
WTC
1
1008
904
904
104-106
86-89
0.17
0.23
I
1
C80-F1-1
WTC
1
C155-F1-1
WTC
1
HH-Fl-1
WTC
1
C26-F1-1
WTC
1
beam
83-86
605
98-101
0.23
0.23
0.17
0.19
C
0.90
0.87
1.08
1.19
Mn
P
77-80
603
92-95
0.23
0.73
904
107
1
Mn
0.74
1.06
P
<0.005
0.006
0.027
0.008
<0.005
<0.005
<0.005
S
0.014
0.015
0.016
0.008
0.017
0.014
0.015
S
Si
0.10
0.02
0.03
0.03
0.03
0.03
0.03
Si
Ni
0.05
0.02
0.02
0.01
0.02
0.02
0.04
Ni
Cr
0.04
0.02
0.02
0.02
0.03
0.02
0.03
Cr
Mo
<0.01
0.01
0.04
0.01
<0.01
<0.01
<0.01
Mo
0.05
0.08
U.Uj
U.Uo
U./4
U.uz
<0.005
< 0.005
<0.005
<0.005
0.065
0.059
Cu
0.24
\']
u
_1
L
0.036
V
I
V/N
Nb
<0.005
<0.005
< 0.005
<0.005
<0.005
<0.005
<0.005
Nb
Ti
<0.005
<0.005
< 0.005
<0.005
<0.005
<0.005
<0.005
Ti
Zr
<0.005
<0.005
< 0.005
<0.005
<0.005
<0.005
<0.005
Zr
Al
<0.005
<0.005
< 0.005
<0.005
<0.005
<0.005
0.018
Al
B
<0.0005
<0.0005
< 0.0005
<0.0005
<0.0005
<0.0005
<0.0005
B
N
0.0067
0.0041
0.003
0.004
0.0049
0.0099
0.0065
N
6.6
9.1
ratio
specified
5.4
36
36
36
36
36
42
50
41.9
31.1
33.2
n/d
42.2
n/d
50.3
YP
NYP
YP
NYP
YP
YP
NYP
39.6
32.4
53.4
34.4
50.9
54.1
n/d
F, (ksi)
F^
avg
(web)
(ksi)
Yield
behavior''
Fy avg
(flange) (ksi)
NISTNCSTAR
1-3D. WTCInvestigation
49
Chapter 3
Shapes
Type
^.
WTC
WTC 2
WTC
1008
104-106
904
904
86-89
1
^
C155-ri-l
C80-F1-1
WTC
1
WTC
1
HH-Fl-1
WTC
1
C26-F1-1
WTC
1
beam
83-86
14WF184
12WF161
12WF92
27WF94
77-80
12WF190
904
107
12WF161
1.090
0.905
1.060
0.840
0.905
0.545
0.576
1.748
1.486
1.736
1.378
1.486
0.865
0.872
(in.)(b)
1
605
98-101
603
92-95
14WF237
//(in.)(b)
a.
C71-FL-2
C65-F1-1
C30-F1-1
Yield behavior
YP = yield
point behavior exists in stress-strain curve
NYP = no yield point behavior exists
b. tf
=
Key:
flange thickness;
/„
= web
in stress-strain
curve
thickness.
n/d, not determined.
Note: [Highlighted items are particularly relevant for
1.8
1
r
T
1
1
1
1
1
1
1
1
1
F = 36
r
steel identification.!
ksi
r
1
1
1
specimens
-|
1
1
r
offset for clarity
Core Columns
1
L orientation only
1.6
^plates
D
owide-flange webs
'o
Q)
1.4
values <
Q.
CO
D
CD
1
arise from
loss of yield point
1.2
W
cd
0
1.0
-obars represent
max and min
points without bars represent single
values
measurements
0.8
35
40
45
Specified F
50
55
ksi
Figure 3-19. Ratio of measured yield strength to specified minimum yield strength for
longitudinal tests of core column steels.
50
NISTNCSTAR
1-3D,
all
WTC Investigation
Room-Temperature Tensile Properties
Summary
Table 3-9.
of mechanical properties and chemical compositions for steels
from core box columns.
Specimen B61522-FL-3
B61521-FL-l
Fuji mill
C88C-
C88A-
C88B-
C88B-
C88B-
test
F2-1
Fl-1
Fl-1
F2-1
F3-1
report^
1
WTC
nr^tion
1
WTC 2 WTC 2 WTC 2 WTC 2 WTC 2
WTC 2
504
33-36
15-18
c
0.17
0.16
Mn
0.81
P
TT
803
801
77-80
801
77-80
0.20
0.18
0.19
0.15
0.18
0.18
0.98
0.96
0.98
1.15
1.11
0.86
0.87
<0.005
0.024
0.013
0.029
0.013
<0.005
<0.005
<0.005
S
0.011
0.013
0.008
0.020
0.021
0.011
0.014
0.016
Si
0.20
0.24
0.20
0.04
0.05
0.09
0.03
0.03
Ni
0.02
0.01
0.02
0.02
0.02
0.02
0.02
Cr
0.03
0.02
0.02
0.03
0.01
0.01
0.02
Mo
<0.01
<0.01
0.06
0.02
<0.01
<0.01
<0.01
Cu
0.05
0.05
0.05
0.05
0.02
0.02
0.03
V
<0.005
< 0.005
< 0.005
<0.005
<0.005
<0.005
<0.005
Nb
<0.005
< 0.005
< 0.005
<0.005
0.030
0.011
0.013
Ti
<0.005
< 0.005
< 0.005
<0.005
<0.005
<0.005
<0.005
Zr
<0.005
< 0.005
< 0.005
<0.005
<0.005
<0.005
<0.005
0.031
< 0.005
<0.005
<0.005
<0.005
<0.005
< 0.0005
0.0024
<0.0005
<0.0005
<0.0005
0.005
0.004
0.006
0.004
0.006
0.013
N
specified F, (ksi)
F, avg (ksi)
Yield behavior''
Column Type
Yield behavior:
=
Key:
36
36
42
42
42
42
42
40.4
38.8
38.4
49.7
n/d
n/d
n/d
n/d
YP
YP
n/d
NYP
n/d
n/d
n/d
n/d
n/a
Type 378
n/a
n/a
n/a
n/a
1.5
1.5
1.44
2.06
by 453.75
3.0
in.
tested
Aug
5 1969, rolled at Fuji Hirohata
64,9 ksi; total elongation measured on gage length of 8
YP = yield
/
36
lb plate 65.5 in,
point behavior exists in stress-strain cur\
NYP = no yield point
c.
0.0070
1.88
/(in.r
TS=
0.0100
Type 354 Type 380
b.
A.
801
77-80
B <0.0005 < 0.0005
Data for 25.263
\_
801
80-83
Al
a.
1
801
80-83
behavior exists
in stress-strain
in:
=
El,
= 32
1.44
Works,
percent,
e.
curve.
plate thickness.
n/a, not applicable; n/d, not determined.
NISTNCSTAR
1-3D,
WTC
Investigation
51
Chapter 3
Figure 3-20. Yield behavior in tests of webs of four wide-flange
core columns specified as Fy = 36 ksi.
Fy = 36 ksi Wide-Flange C-65
= 36
12WF161 shape from
WTC
Specimen C-65
is
This location
on the opposite side of and well-below the impact zone. None of the eight specimens
tested, six
is
from
a flange
elongation. There
is
a F,
is
1
column
line
ksi.
The
904 between
floors 86-89.
and two from the web, exhibit either yield point or significant yield point
a small yield point elongation in the
two specimens from
only one tenth as long as the usual (1-3) percent. Only one of the eight
above 36
the
tests
web specimens is deformed
specimens were harvested away from the large tear
section that
Appendix B, although the
is
ksi
is
the source of the
the source of specimens from the flange, see
web
(Fig. 3-20), but
it
produced a yield strength
into an
in the
"S" shape, see
web. The section that
Appendix B, shows no evidence of significant
deformation, but the lack of yield point and yield point elongation indicate that the material has more than
several percent prior strain.
There are several possible explanations
behavior.
shapes as
flange shapes from slightly after the
AISI
for the
low yield strength, given the absence of yield point
Galambos (1978) reported the mean value of yield strength in mill test reports for English A 36
1 .22x36 ksi = 44 ksi, which is consistent with data in Barsom (1988) on U.S. heavy A 36 wide-
WTC construction era. For such a mean large excess strength, the
variability study (1974) predicts virtually
no possibility of finding under- strength product
unlike the estimate for high-strength perimeter column plates in the previous section.
strengths of the other
than this
mean
value.
much
rolling,
One
limitation of the
AISI
variability study is that
ASTM Group
1
97 percent of the members of its
and 2 shapes, while the
WTC core columns tested are
much less reduction during
average strengths may be much closer to
heavier Group 3 and 4 shapes. Because these heavier shapes undergo
and cool more slowly from the rolling temperature, their
the specified
52
yield
WTC wide-flange specimens tested that exhibit yield point phenomena are all less
data set for wide-flange shapes are
all
tests,
The measured
minimum
than the heavily rolled, fast-cooling lighter shapes in the AISI study.
NISTNCSTAR
1-3D,
It is
also
WTC Investigation
Room-Temperature Tensile Properties
possible that this difference represents different steelmaking practice for the wide-flange columns, which
probably came from Yawata Iron and
Steel.
There are several other possible origins for the low value of the measured yield strength. The typical
YP and
difference between yield point
shortfall,
mean of the two
although the
could account for an additional 2
tensile tests, the
yield strength
ksi. Finally,
YS or F,.
only 30.5
tests is
were the prior
(3 ksi or
testing rate in the
strain in the opposite direction
of the
for part
NIST
tests
from the
Bauschinger effect (Uko 1980, Tipper 1952, Elliot 2004) could further reduce the
measured yield strength (10-20) percent from the expected value. In
3 percent strain after pre-straining in the opposite direction
the difference decreases with increasing strain.
= 36
strength above F,
very similar to the other F,
stress is similar to
is
much
particular, the
less
The sum of these three
flow stress in the
first
than in the forward direction, but
effects could raise the yield
ksi.
Although the yield strength of column C-65
is
more) could account
The slower
ksi.
= 36
ksi
is
less than the specified
columns. Fig. 3-20.
column C-30. Furthermore,
minimum,
By approximately
the tensile strength
and
the stress-strain behavior
3 percent strain, the
total elongation
measured
flow
in
specimens from C-65 are consistent with the requirements of A 36.
The
chemistr>' of
C-65 meets
ASTM A 36.
Fy = 36 ksi Wide-Flange C-71
12WF190 shape from WTC 1 column line 904 between floors 77-80 specified to
have F, = 36 ksi. Both web specimens exhibit yield point and yield point elongation behavior, but the
measured yield point is less than 36 ksi; the mean value is YP = 33.2 ksi. In contrast the flange specimens
Specimen C-71
a
is
exhibit significant evidence of substantial prior deformation: F,
elongation.
The measured chemistry,
total elongation,
and
= 53.3
ksi
and absence of yield point
tensile strength are all consistent with the
requirements of A 36.
Arguments made
low measured yield strengths of specimen C-65 apply
for explaining the
to
specimen
C-71 as well.
It is
likely that the high strength
and lack of yield point behavior of the specimens from the flanges arose
from bending of the column about the web
deformed columns and estimates the
deformation of the column
is
axis.
Appendix B summarizes an analysis of the shape of the
strain in the flanges
from the bending of the column. The measured
consistent with the elevated strength of the flange specimens,
from the portion of the flange with
which came
tensile strain.
Fy = 36 ksi Wide-Flange C-80
Specimen C-80
have F, = 36
specified
is
ksi.
a
14WF184 shape from
The
minimum:
WTC
1
column
line
603 between floors 92-95 specified
yield strength of the single flange specimen tested
F,
=
34.4
ksi.
The
is
slightly
to
lower than the
stress-strain curve does not exhibit yield point or yield point
elongation. Reintroducing the expected yield point behavior to the
measured yield strength would
probably increase the yield strength enough to be consistent with the requirements of A 36. The
chemistry, tensile strength, and total elongation of C-80 are consistent with the requirements of
ASTM
A 36.
NISTNCSTAR
1-3D.
WTC
Investigation
53
Chapter 3
Fy = 36 ksi Wide-Flange C-1 55
14WF184 shape from WTC column line 603 between floors 92-95 specified to
web specimens exhibit yield point and yield point elongation behavior, and the
three are well above 36 ksi; the average yield point is YP — 42.2 ksi. The chemistry,
Specimen C-155
is
have Fy = 36
All three
ksi.
yield points of all
tensile strength,
a
and
1
elongation
total
all
A
meet
36.
Like specimens from C-71, the flange specimens show evidence of prior deformation: elevated yield
strength,
and lack of yield point and yield point elongation. Like column C-71, the magnitude of the
elevated strength and lack of yield point in the flanges
is
consistent with the
measured defomiation
column. The tensile specimens came from the portion of the flange that had been strained
Appendix B
details the calculation.
HH
Fy = 42 ksi Wide-Flange
Specimen
Fy = 42
HH
ksi.
in the
in tension.
is
12WF92 shape from
a
WTC
column
1
605 between floors 98-101 specified to have
line
The two longitudinally oriented specimens from
YP =
elongation behavior with
54.1 ksi.
Although the mill
and yield point
the flange exhibit yield point
test report requires
given that the flanges usually have lower strength than the webs,
data from a
likely that a
it is
web
web specimen,
test
from specimen
HH would have produced a yield point YP > 42 ksi. The chemistry of specimen HH is consistent with the
requirements of both ASTM A 441 and A 440 with its Cu content and A 572 types 2 and 4 because of its
high vanadium content. The chemistry of HH-FL-1
Both
ASTM A 441
A 440 require
or
is
inconsistent with
a group 2 shape like
HH to have
measured yield point of HH-FL-1. The measured yield point
A 572
Grades 42, 45, 50, and 55, but
contract.
A
However,
572 Type 4
572 does not appear
documents (Betts 1967) indicate
rolled shapes in the core. Other
an exception to supply
historical
A
written to be satisfied
that
have been supplied
It is
documents (Monti 1967) indicate
= 50
to F,
most
likely that
is
lacks Nb.
it
less than the
of approved
steels in the
PONYA
Bethlehem V-series of steels, which
the
supplied in 42 ksi and 45 ksi grades.
steel for the
Yawata requested and was granted
HH-FL-1 does not meet. The
supply all rolled shapes with F,. > 36 ksi
that
composition, which
documents do not indicate whether Yawata intended
to this specification or not.
which
ksi,
Yawata supplied about half of the mass of the
own "A 441 modified"
its
YP = 50
in the list
by
Ex-Ten, because
also consistent with the requirements of
PONYA contract and presumably could be
were permitted under the
Historical
is
is
USS
HH-FL-1
to
represents a
Yawata
A 441
shape, which
would
ksi.
Fy = 50 ksi Wide-Flange C-26
Specimen C-26
is
a
The measured yield
1
ksi.
None of the
27WF94
in,
WTC
1
107th floor specified to have F,
web specimens
are less than 50 ksi, but both
is
quite deformed. Indeed, the
flat,
is
consistent with the requirements of
vanadium and nitrogen
A 44
content.
Because
it
The web
full-thickness, 8 in. tensile
so the absence of a yield point and significant yield point elongation
the requirements of
54
in the
stress-strain curves exhibit yield point or yield point elongation.
The chemistry of C-26
the
beam
strengths of two of the three
source of the specimens
about 1/8
shape from a
A
is
= 50
by
that
ksi.
less than
is
the
specimens bowed
not surprising.
572 Type 2 and Type 4 as well, based on
lacks of sufficient Cu, the chemistry
is
inconsistent with
1
NISTNCSTAR
1-3D.
WTC Investigation
Room-Temperature Tensile Properties
Given the yield strength and chemistry C-26 could be
a
Bethlehem V-series shape, which the
PONYA
steel contract did allow.
Fy = 36 ksi
Box B6152-1
Specimen B6 152-1
is
a 2.06 in. thick flange plate
from
803 betw een floors 15-18 specified to have Fy = 36
ksi.
380 box column from
a type
The two
WTC
stress-strain curves are
only one exhibits a definite yield point. The measured yield strengths are both F, > 36
The chemistry
is
consistent with the requirements of
A 36 plate
a Fuji steel
column
1
line
very similar, but
ksi,
however.
ASTM A 36 and is rather similar to the chemistry of
(Morris 1969) shipped to Stanray Pacific, Table 3-8. That similarity and
its
thickness (2.06 in) are consistent with the historical documents (Warner 1967) that Fuji Steel supplied
most of the plates with
Fy = 36 ksi
?
<
3 in. in the Stanray Pacific contract for the
welded core columns.
Box B6152-2
Specimen B6 152-2
is
a 1.88 in. thick flange plate
504 between floors 33-36 specified
from
have F, = 36
to
ksi.
354 box column from
a type
WTC
B6152-1-F1 and consistent with
similar to
is
column
line
Both longitudinal specimens exhibit a yield point
and yield point elongation. The yield points exceed the requirements of A 36: YP = 40.4
chemistry of B6152-2-F1-2
1
A
The
ksi.
36.
Fy = 42 ksi Box C-88c
Specimen C-88c
a 1.5
is
in.
thick flange plate from a type 378 box
between floors 80-83 specified
have F, = 42
to
between C-88a and C-88b
the junction
the single longitudinal
7S = 49.7
test,
is
ksi. Originally,
the intact
ksi, (0.2
welded
it
column from
was
WTC 2 column line 801
part of C-88a, Fig. 3-21.
splice at the 80th floor.
percent offset)
is
much
The
Note
that
yield strength of
larger than the F,.
= 42
ksi yield
strength requirement.
The chemistry of C-88c
the
anomalous. Table 3-9.
PONYA steel contract.
Because
the
is
A
it
Its
does not contain
lack of vanadium
Nb
either,
572 specification. Although the
it
It is
inconsistent with
makes
it
all
ASTM
inconsistent with the requirements of
inconsistent with the chemistry requirements of
PONYA
steel contracts
do not mention
A
evidence (Marubeni-Iida 1967) of shipments of A 572 steels to Stanray Pacific.
inconsistent with the requirements of
A 440.
There are several
notably proprietary Lukens grades, for which
possibly
it
?
<
1
572, there
Its
steels listed in the
A 441.
the variants of
is
archival
low Cu content
is
PONYA contract, most
cannot locate typical or specified compositions, so
minimum
compositions, and are thus of limited
Contemporaneous documents (Warner 1967) indicate
.75 in.
with F,
NIST
all
could have been shipped to one of those specifications. Summaries such as Woldman's (1990)
indicate only specified
alloys.
specifications hsted in
The thickness of the C-88
> 42
ksi for the
are specified as F,
= 42
their general contract.
utility in identifying specific
that Fuji steel supplied all plates
plates (1.5 in.) identifies Fuji as a possible source.
box columns were rather uncommon; only 69 of the
ksi.
The
Because of their
PANYNJ
rarity,
1
with
However,
plates
202 columns per building
Stanray Pacific could have purchased these outside
documents reviewed by NIST yielded no correspondence on
it is certainly possible that the PANYNJ documents do not
substitutions or approvals of such a plate, but
contain a complete record of the construction correspondence.
mills might
have supplied
NISTNCSTAR
1-3D.
NIST was
not been able to identify which
this plate.
WTC Investigation
55
Chapter 3
The other
F,
= 42
ksi steels
similar, but not identical.
and contains B, but
much
floors below, contain
Nb;
from the adjacent
plate
and column respectively, C-88a, are chemically
Table 3-9. The chemistry of the flange opposite C-88c (C-88a-FI)
less
Mo. The
which are
plates of C-88b,
slightly thinner
their chemistries are consistent with the requirements
Circled locations denote approximate
specimen locations
is
different
and originate from the
of A 572, grade 42.
for
chemical analysis
Source: NISI.
Figure 3-21. Section of C-88c that
is
the source of the specimens.
Trusses
Contemporaneous construction documents indicate
that
Laclede (1968b, 1968c) supplied the
steel in the
A 36 and A 242. Although modem versions of A 242 specify
requirements for corrosion resistance, A 242-66 left the buyer and seller to set these requirements. A 36
specifies a minimum yield point, F, = 36 ksi, while A 242 specifies a higher yield point, F, = 50 ksi.
floor trusses to
two
specifications:
Table 3-10 summarizes the dimensions of the main truss components, the specifications they were
supplied
to,
and where they were corrmionly used. Figure 3-22 plots the
strength or yield point to the specified
56
minimum
value for
ratio
of the NIST-measured yield
all truss steels tested.
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
The chemistries of all of the
A 242.
and
Their strengths
truss
all
exceed the specified minima. Table 3-1
mechanical property data for the truss
summarizes the chemistry and
truss
component dimensions and standards.
Shape
Specification
A
36
3
A
36
1.09
A 36
A 242
1
in.
1
13/16
Lower chord of many common 60
2
in.
Main web of many 60
round bars
in.
X 1.5
ft
ft
trusses
ft
trusses
tmsses
Unknown
round bars
in.
I^pH F'or
1
X 2 in. bulb angle
in.
Upper chord of many common 60
bulb angle
Upper and lower chord of many common 36
A
242
0.75 in round bars
A
242
0.92
in.
0.98
in.
1.14
36
steels.
Common
Ta ble 3-10.
A
components tested are consistent with the requirements of both
Main web of bridging
in.
trusses
trusses
Main web of many common 36
round bars
ft
ft
trusses
Truss Angles
Four of the seven 2
A
572
steel.
in.
x
1
.5 in.
The measured
were supplied
to
A
36.
bulb angles have added vanadium, which makes them similar to a
yield strengths of all of the truss angles tested are
which
is
consistent with the requirements of both
strength of one of the three bulb angles specified as
Truss
The
the
A
all Fy > 50
36 and
A 36 exceeds the required
ksi,
A 242.
even
The
maximum by
1
modem
if
they
tensile
ksi.
Round Bars
yield strength of an
D=
1.09
in.
A
36 web round bar with
D^
1
.09 in.
is
Fy = 47.4
diameter bars are consistent with the requirements of
mechanical properties of all round bars originally specified as
A
ksi.
A 36.
The chemistries of both
The chemistries and
242 are consistent with the requirements
of that standard.
NISTNCSTAR
1-3D,
WTC Investigation
57
Chapter 3
Table 3-11.
Summary
of mechanical properties and chemical compositions, and
specifications for truss steels tested.
Truss Bulb Angles
Tl-BA-
C137f- T-l-TA-
C53TA-1
C53-
C132-
C137a-
C53-
Plate""
TA-2
TA-3
TA-5
Plate"
TA-3
BA3
c
0.20
0.22
0.18
0.19
0.18
0.18
0.19
0.17
Mn
0.77
0.46
0.84
0.82
0.84
0.86
0.90
0.60
0.024
0.010
0.020
0.016
0.021
0.013
0.040
p
0.009
s
0.032
0.028
0.024
0.029
0.020
0.028
0.029
Si
0.06
0.15
0.04
0.07
0.07
0.04
0.06
0.02
Ni
0.09
0.10
0.08
0.08
0.06
0.08
0.08
0.09
0.10
0.05
0.12
0 08
0
V/. 08
V/O
0.01
0.02
0.05
0 01
0.22
0.29
0 04.4
0.30
0
04^J
U.UH
U.U J 7
u.uuy
^ U.UUj
^ U.UUj
^ U.UUJ
Cr
0.
0.09
12
Mo
<0.01
Cu
0.26
<
0.
<
0.01
0.48
VV
<
fi
on^
<rn
City's
u.uuj
*^
<
0.01
0.01
0.28
0.32
^ U.UUJ
U.UJO
U.U 1 o
^ U.uUj
^ U.UUj
^ U.UUJ
^ u.uuj
l\0
Ti
1 1
10
^ U.UUj
U.U 1
1
Al
0.045
< 0.005
0.028
0.032
B
N
<0.0005
<0.0005
<0.0005
<0.0005
dimensions
1.5 in. X
2
1.25
1.5 in.
0.0080
0.0080
in.
in.
0.25
A
Spec
Fv(ksi)
A
242
X
X
in.
242
n/d
n/d
0.0010
2
in.
0.0080
X
2
in.
X
1.5 in. X
1.5 in.
0.25 m.
0.25
A
A
242
X
in.
242
Truss
\J.U\JJ
^ U.UUj
^ U.UUj
^ \j .\}\} J
^u.uuj
<n
DOS
^U.UUJ
^ U.UUj
^ U.UUj
^ U.UUJ
0.038
0.015
0.029
0.32
n/d
<0.0005 <0.0005 < 0.0005
0.0100
2
58.8
n/d
^
in.
X
1.5 in.
0.25
0.0066
2 in.x
X
in.
A 242
0.25
0.0160
2
1.5 in.x
in.
A 242
1
in.
X
.5 in.
0.37
A
0.0090
2
X
in.
242
^2 2
n/d
C53-
M32-
LR-r
LR-1
C106SR-1
0.20
0.86
0.64
n/d
< 0.0005
in.
X
1.5 in.
0.25
A
X
m
3
in.
2
in.
0.37
A
242
55.5
X
X
in.
36
57.2
Round Bars
A242-
€53Tl-SR-1
MR-1
0.20
0.84
0.22
P
0.008
S
Si
C
C137aSR-3 Tl-LR-l**
66
A
36-66
0.21
0.92
0.24
0.99
0.79
0.18
0.79
0.018
0.007
0.008
0.018
0.020
0.011
0.04
0.027
0.031
0.025
0.019
0.030
0.032
0.023
0.05
0.07
0.08
0.08
0.05
0.06
0.07
0.05
Ni
Cr
0.05
0.07
0.08
0.04
0.03
0.08
0.06
0.07
0.05
0.11
0.04
0.10
0.06
0.09
Mo
<0.01
0.03
< 0.01
<0.01
0.01
0.03
0.01
Cu
0.12
0.22
0.27
0.08
0.27
0.04
0.035
0.041
0.038
0.038
Mn
V
Nb
Ti
Zr
Al
B
N
diameter
Spec.
Fv(ksi)
<0.005
<0.005
<0.005
0.044
<0.0005
0.0083
0.75 m.
A 242
< 0.005
<0.005
< 0.005
<0.005
< 0.005
<0.005
0.032
0.033
<0.0005 < 0.0005
0.0100
0.0110
0.92 m.
0.92 in.
A 242
A 242
n/d
n/d
<0.005
<0.005
<0.005
0.033
<0.0005
0.0080
0.92 m.
A 242
60.0
n/d
<0.005
<0.005
<0.005
<0.005
<0.005
<0.005
<0.005
<0.005
<0.005
0.033
<0.0005 <0.0005
0.0110
0.0100
1.09 in.
1.09 m.
A 36
A 36
47.4
a.
From
b.
Thickness approximately 0.23
c.
Originally attached to perimeter column N8. Chemistry not determined.
d.
Originally assigned as
n/d
0.21
0.26
Type
V
0.22
1.25
0.05
0.26
0.031
<0.005
<0.005
<0.005
0.017
<0.0005
0.0110
A
n/d
36
36
A
242
50
a bridging truss.
D=
1
in.
.0 in.
e.
Yield point reported. Other alloying elements must be reported.
Key: n/d not determined
58
NISTNCSTAR
1-3D,
WTC Investigation
—
—
—
r
Room-Temperature Tensile Properties
1.8
———— ————
— — — — — — ——— —= 50— —specimens
I
1
I
I
I
I
I
I
I
T
I
I
I
I
I
I
I
I
\
I
I
I
]
\
Fy
offset for clarity
ksi
r
1
1.6
D
g
Q
1.4
Q.
CO
LJ_
D
1.2
(D
I
O)
(0
Truss Components
0
L orientation only
1.0
bars represent
J
0.8
I
I
I
I
I
I
I
I
40
35
max and min
points without bars represent single
L
I
I
J
I
45
50
Specified F
values
measurements
I
I
L
60
55
ksi
Figure 3-22. Ratio of measured yield strength to specified
yield strength for all longitudinal tests of truss steels.
minimum
Truss Seats
The
structural engineering plans specify that all truss seats
Fy = 36
steel
ksi.
to a specific
ASTM
minimum
to
have a minimum yield strength
steel.
The plans do not
3-23 plots the
ratio
all truss seats
characterized.
terms of averages and ranges because the mechanical data
in
require that the
of the NIST-measured
for all longitudinal tests of truss seat steels.
Table 3-12 summarizes the chemistry and yield data for
summarized
36
standard, however. Figure
yield strength or yield point to the specified
are
A
Presumably, the fabricators would have used
conform
were
The chemistry
come from one
set
data
of
specimens and the chemistry data come from another. The chemistries of all the perimeter truss seats
characterized are consistent with the requirements of
A 36
is
A
for yield and tensile strength and total elongation.
slightly less than
slightly
low value
A
is
36 allows:
ksi
-
58.6 ksi
maximum that A
the high strength
NISTNCSTAR
is
1-3D.
= 34
not remarkable.
stronger than required: F,
TS = 80
F,.
ksi.
36. All the perimeter
The
column
truss seats
meet
yield point of the core truss seats (C128-T1)
Because of the natural
variability in product tests, this
One of the other core column truss seats tested is substantially
and TS - 82.2 ksi. The tensile strength slightly exceeds the
36 allows. This truss seat
is
strengthened with vanadium, see Table 3-12;
consistent with the chemistry.
WTC
Investigation
59
Chapter 3
1.8
T
1
1
\
1
1
1
]
36
ksi
r
1
1
1
Fy =
1
specimens
offset for clarity
n
'
!
Truss seats
L orientation only
T3
0)
Q
1.4
Q.
CO
D
1.2
CD
03
CO
0
^
1.0
bars represent
max and min
points without bars represent single
J
0.8
32
I
I
34
I
I
I
I
I
I
40
38
36
Specified
-
measurements
I
I
values
F^, ksi
Figure 3-23. Ratio of measured yield strength to specified minimum yield strength for
longitudinal tests of truss seat steels.
Bolts
3.4.2
the requirements of their respective standards.
M-2
3.4.3
is
and
consistent with the requirements of the specification for
A 325
Type
1.
Welds
One problem
fabrication
A 325
A 490 bolts. The measured strengths exceed
The chemistry of an A 325 bolt recovered from perimeter
Table 3-1 summarizes the results of tests of the
panel
al
in
comparing weld properties
to those reported in fabrication
documents usually reference welding procedure
data.
documents
test
specimens
(tensile,
that the
These data are taken from
with very large gaps between the base plates. The advantage of this large gap
remove mechanical
is
is
that
it
test
makes
it
welds
easy to
impact and bend) that are purely weld metal, and so
accurately reflect the performance of the weld metal itself
The disadvantage
is
that this large
gap does
not simulate the conditions of the production weld in the component, where dilution of the weld with
some of the base metal
will
change
its
properties.
carbon content in the base metal, which
is
The change
is
usually to higher strength since the
higher than that of the
filler
metal, increases the hardenability
of the weld composition.
60
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
Table 3-12.
Summary
of mechanical properties, chemical compositions for truss seat
steels tested.
Perimeter Seat ChemistrA
M34-TSA-L5
C128-T1
%
%
0.16
0.24
0.19
0.47
0.41
1.16
0.83
0.006
0.008
<0.005
0.013
0.01
S
0.022
0.030
0.015
0.006
0.005
Si
0.06
0.07
0.06
0.05
0.038
Ni
0.09
0.16
0.06
0.022
0.014
Cr
0.07
0.10
0.05
0.053
0.018
Mo
<0.01
<0.01
0.035
0.02
Cu
0.24
0.32
0.27
0.027
V
<0.005
<0.005
0.048
0.002
Nb
(a)
0.046
0.002
n/d
Ti
<0.005
<0.005
0.002
n/d
Zr
<0.005
<0.005
n/d
n/d
Al
(b)
0.034
n/d
n/d
B
<0.0005
<0.0005
n/d
n/d
N
0.0090
0.0280
n/d
n/d
Average
Max
Min
%
%
%
c
0.19
0.24
Mn
0.44
P
Description
N8-C1B1
<0.005
<0.005
0.0040
Perimeter
Perimeter
Perimeter
Perimeter
seat
seat
seat
seat'
(ksi)
Yield behavior
b.
c.
Longitudinal orientation.
d.
Transverse orientation.
e.
Two
Key: n
with [Nb] =0.046 percent,
with [Al]=0.034 percent,
all
all
Perimeter
Perimeter
seat*"
seat**
Core
seat
Core
seat
'
A
One specimen
One specimen
N99-C1B
0.20
Specification
a.
N9-C1T1
36
A
36
A 36
A
36
A
36
40.7
40.9
41.1
58.6
34.0
YP
YP
YP
YP
YP
others <0.005.
others
<
0.005.
of three specimens have YP.
d,
not determined;
NYP; no
yield point behavior;
YP. yield point behavior
in stress-strain curve.
Notes: Chemistry averages based on 7 specimens of perimeter truss seats not characterized for strength. Chemistry expressed in
mass
fraction.
Welds
in
the Perimeter
Columns
The perimeter column panels were
large
and complicated, so a variety of welding processes were used
during the various steps in their fabrication. Although welding procedures were sometimes changed
during the construction
(SHCR
1968), the floors that the aircraft struck appear to have been constructed
using gas-shielded flux-core electrodes to
make
the butt welds that joined the spandrels to the inner
web
of the panels, shielded-metal-arc electrodes for tacking the diaphragm plates inside the columns, tandem
submerged arc to make the long longitudinal fillet welds that formed the box shape, and a combination of
flux-core and shielded-metal-arc electrodes to finish the ends
NISTNCSTAR
1-3D,
WTC
Investigation
(PCF
1967).
61
Chapter 3
Pacific
that
52
Car and Foundry requested permission
produced Fy = 84.5
ft
lbf at 0 °F,
ksi,
TS = 94.5
when welded with
(SHCR
ksi, total
1968) to change to a different flux-cored electrode
elongation EI, = 24 percent, and impact toughness of
14 passes on 0.75
in.
thick plate. Apparently, they wished to use this
electrode for the butt welds that connected the spandrels to the inner web.
procedure for the columns
It is
likely that they
used
this
of impact. Pacific Car and Foundry's welding procedure (PCF
at the level
1967) specifies a series of weld strength ranges that correspond to the base metal strength ranges, and
typical properties of the flux-cored electrodes to be used.
Similarly, the submerged-arc fillet welds used
=
two types of flux, one with higher alloying
levels, to
TS =
produce two strength levels: F,
81 ksi,
92.3 ksi, and F, = 105 ksi, 75 = 1 18 ksi (PCF 1967). The
lower-strength flux would have been specified for the F,. = 55 ksi (nominal) strength of perimeter column
N-9 (WTC 1, column 154). These procedure data compare well with the measured strengths for a fillet
weld in N-9 ofF,, = 85.1 ksi, TS =
ksi, and F,- 84.8 ksi, TS^ 103.3 ksi. The strengths are slightly
higher than those listed in the procedure document (F,. = 81 ksi, TS = 92.3 ksi), probably because of
m
dilution of the
weld by the higher-carbon base metal.
Tests of the transverse strength of the perimeter columns at the region of the
fillet
welds
in
column N-8,
Table 3-3, show that the weld region was stronger than the adjacent plate material. With the weld metal
milled from the surface of the flange, the failure occurred outside the weld region. Without the weld
removed, the failure occurred
at the
weld
root, indicating that residual stress or geometrical effects
explain the rips that occurred adjacent to the longitudinal welds in
many of the
recovered columns.
Together, these transverse data provide a measure of the strength and ductility of the longitudinal
welds
in the
fillet
columns. Since none of these transverse specimens actually broke in the weld, the results are
not a true measure of the weld strength, but rather of the weakest region across the entire joint (base metal
to
weld
The
to base metal).
failure location,
which
is
adjacent to the weld, matches the general
observations on the recovered sections, where rips progressed through the regions adjacent to the welds,
but seldom through the welds.
Welds
in
the Core
The longitudinal
fillet
Columns
welds used
smaller than the plates.
thick
A 36 plate to
at failure for
to fabricate the core
An example
a 1.25 in. thick
box columns from
(Stanray Pacific 1967)
A 36 plate using
two transversely loaded
fillet
weld
a
tests
is
a nominal 5/8
tandem submerged
was 42,400
linear inch in
two
tests.
property data on a 0.5
plate to fabricate a
fillet
welds of 25,600
Another document, (Stanray
in.
triple-wire submerged-arc
used
in. fillet
and 84.9
much
to join a
weld used
lb per linear inch
to join a 1.5 in.
failed at
A
Stanray Pacific also
and 26,200 Ibf per
weld
36 plate to a 1.25
236,000
in.
The equivalent shear
ksi.
Pacific, undated) provides transverse fillet
column. Just over 5 linear inches of this weld
1
The average shear load
arc weld.
Ibf per linear inch.
strengths, calculated using the actual cross-sectional areas, are 82.9 ksi
measured longitudinal shear loads for these
plates tended to be
Ibf, for
in.
a shear load of
45,900 Ibf per linear inch.
Because submerged arc welding was not suitable for the ends of the columns or
align plates during construction. Stanray Pacific (1967b) used a self-shielded
locations, with
E7018 used
was developed on
62
as a root pass
when
for the tack
welds used to
FCAW electrode for these
full penetration welds were required.
The weld procedure
a 1.25 in-thick plate with a 20 degree bevel on a butt weld, and a root gap of 0.5
NISTNCSTAR
1-3D,
WTC
in.
Investigation
Room-Temperature Tensile Properties
Two transverse
tensile tests
75=81
ksi.
Welds
in
Truss Tl
gave tensile strengths TS = 84.4 ksi
The wide gap
(failure in the weld).
(failure in the
HAZ) and TS =
84.25 ksi
also permitted determination of the all-weld-metal tensile strength
the Trusses
built
is
from 2
in.
x 1.5 in. x 0.25 in. thick bulb angle with a 0.92 in. diameter rod.
The
manufacturer, Laclede (1968) designated this resistance weld assembly as type R-21. Unfortunately there
are
no strength data for
this
assembly type, but
it is
very similar to several others for which data do
Extrapolating from the data for R-12 (56,900 Ibf) and R-17 (56,800 Ibf) suggests an average load
56,700 Ibf for the R-21 assembly (per the manufacturers shear
failure of
was based on loading of an
entire
The shear load of 104.000 Ibf in
and
The
the test of the section
different test
There were
a
from
Some
Tl
truss
is
strengthening
significantly higher than the
may have
occurred by aging during
methods between the manufacturer, who
with a slightly different loading geometry, and
measured load
at
procedure). This procedure
intact truss.
56,700 Ibf suggested by the manufacturer's data.
the lifetime of the truss.
test
exist.
NIST
is
tested an entire truss
another possible source of the difference, but the
higher than expected.
at failure is still
few non-resistance welds
in the trusses, primarily fillet
welds near the ends, but the weld
However, documents (Laclede 1969) provide weld
strength data from procedure and welder qualification tests for both shielded-metal-arc and flux-cored-arc
geometry of these welds was not suitable for
electrodes.
(total
They
test.
indicate shear strengths of 6.8 ksi/in to 7.5 ksi/in for about 9
shear load of 60,500 Ibf to 67,000 Ibf) produced with an
about 9 ksi/in for about 9
in.
of a 1/8
in. fillet
weld
(total
in.
of a 1/8
E7018 shielded metal
in. fillet
weld
arc electrode,
and
shear load near 80,000 Ibf) produced with a
fluxed cored electrode.
3.5
MECHANICAL PROPERTY VALUES
3.5.1
Steel
Early in the Investigation, other
strain
behavior of the
estimating
some
many
WTC
Investigation
steels.
Team members
Responding
requested estimates for the stress-
to these requests in a timely
manner necessitated
properties based solely on historical averages, rather than on experimentally determined
properties. Experimental
strain
NIST
measurements on recovered
behavior for perimeter column and truss
however, are based mostly on
steels.
steels
The
formed the basis
for estimating the stress-
stress-strain curves for core
column
steels,
literamre estimates of the properties of WTC-construction era plates
and
shapes.
Estimating the constitutive behavior of the
WTC
steels required
two
steps.
The
first
step
was
to
make
the
best estimate of the yield strength for each grade of steel by combining measured values, literature
estimates of expected properties, and historical mill test reports.
The second
step
was
work-hardening behavior for each grade using stress-strain relations measured from
work -hardening behavior
For step
1
,
model. The
is
not available in the literature for any
two types of correcfion are necessary
first is to
NISTNCSTAR
1-3D,
WTC
to estimate the
WTC
steels,
because
steel.
to estimate the yield strength for input into a constitutive
correct the experimentally derived or mill test report data for artifacts of testing.
WTC
Investigation
63
Chapter 3
These corrections are expressed as factors
minimum yield
specified
for the core
column
case, correction
steels at the
is
to
be added to the measured yield strengths.
available, a second type of correction
is
If only the
necessary. This
was
the case
time that the modeling groups needed the constitutive behavior. In this
must be made for the
minimum. This
specified
minimum
An
strength
correction
of the as-supplied
fact that the yield strength
is
exceeds the
steel
expressed as a multiplicative correction factor
to the specified
yield.
important additive correction to the estimated yield strength
is
necessary because the experimental and
mill test report values are enhanced because the tests are not conducted at zero strain rate (Beedle 1960).
Properly, the yield strength in a mill test report or experimental determination
For modeling building performance, the appropriate strength
Oyd.
strength a,
Starting
such as produced
Rao
et al.
and
static, Oys,
from the
in the aircraft
Oyj, yield strength
dynamic yield
a
strength,
the static, or zero strain rate, yield
enhancements due
From tests on several
16x1 0~s~ >£">2x 10~^s~'
impact, can be calculated.
(1963) estimated that for strain rates,
and dynamic
is
static yield strength, all the strength
is
,
to strain rate effects,
different structural steels,
the difference
between
can be represented by
^v.-^,. =-(3-2 + 1000^)=
where the
(1963) measured the
actuator.
value
is
termed
£
strain rate,
With
is
,
measured
static yield strength
^dynamic, is
s"',
straining the
The
specimen past
of stmctural
is less
than zero. Galambos (1978) used this relation
use in Load and Resistance Factor Design (LRFD). In
steel for
and the
strain rate in
NIST-conducted
tests is
assumed
to
in rolled wide-flange shapes,
additive correction for the variation in yield strength with location
be 3.3x10
it
is
assumed
s"'.
necessary to
from which the
web
test
make an another
specimen
"H" shaped specimen) of the
section (in the "cross bar" of an
rather than the flange. In typical rolled shapes, however, the flange
most of the load-carrying capacity of the column. Because
rolling temperature, and
its
yield strength
is
it is
is
is
taken.
ksi shapes.
It
was not uncommon
for the
rolled shape,
thicker than the web, and accounts
thicker,
generally lower than the web.
Alpsten (1972), have characterized this strength difference between
nominally 36
it
cools
Many
web and
more slowly from
studies,
correction for variation with location,
Anange- is
flange as 2 ksi to 4 ksi for
measured yield strength of a flange
to
be
I ks-i
_
web
_
-
The
K
'^^flange
to
'
also an additive correction factor:
flange
the
summarized by
2 ksi below the specified nominal value reported in the mill test report for samples from the web.
When
to
'*
WTC era, but not currently, ASTM standards specified that the test specimen for the mill test
report be taken from the
for
Rao
and then locking the
static yield strength for this Investigation, the strain rate in mill test reports is
For estimating the yield strength of steel
During the
yield,
in ksi.
difference between the static and dynamic yield strengths,
an additive correction factor, which
to estimate the properties
be 6x10"^
by
and the strengths are measured
the actuator locked, the load drops slightly at nominally constant strain. This constant
the static yield strength, a,,.
computing the
in inverse seconds,
3-1
^ ^
-^"^
•
insufficient experimental or mill test report data are available for a steel grade,
it is
necessary to use
a second, multiplicative correction factor to estimate the yield strength from the historical averages of that
grade. During the 1960s and 70s, several studies (AISI 1973;
Galambos 1976; Alpsten 1972)
characterized the variability in properties of steels supplied to various standards. These studies answered
questions
64
like,
"What
is
the
mean value of the
yield strength of the plates or shapes that meet
NISTNCSTAR
1-3D.
A 36?"
WTC
Investigation
Room-Temperature Tensile Properties
They answered
this question
by examining thousands of mill
product testing. Because the tension
test to certify the
mechanical properties
the production process, scrapping a heat of steel because
undesirable. Thus, steel mills generally strive to
reports, but not
make
it
minimum
specified
is
conducted near the end of
did not meet the intended specification
steels in
specification. Typically, the yield strengths in the mill reports
minimum
by doing independent
is
which the strength exceeds the intended
from the WTC-construction era exceed the
The exact value depended on the specified
was supplied as plates or shapes. Of course, the yield
values by 5 percent to 10 percent.
yield strength, and whether the steel
strength in the mill report will never be less than the standard calls for, because the steel could not have
been sold as meeting the standard. The results of these studies are necessary
of WTC
The
steels
when no
correction, kp,
other corroborating evidence
from specified minimum
For example, for rolled
plates, the
mean
is
the specified
minimum
static yield strength,
o",.^
,
is
a multiplicative factor.
yield strength from the mill test reports, YS,
is
3-3
k^F^.,
yield strength.
Estimating the yield strength of a given
The
available.
to historical average yield strength
yS =
where F,
is
in estimating the properties
WTC steel requires some or all of these tliree correction factors.
for perimeter
columns and
truss steels
was calculated from
the average
value of the NIST-measured yield strengths, YS, corrected for dynamic effects.
a =YS + k,dynamic
vi
where
^dvnamic is
steel, these
calculated from Eq. 3-1.
data were
historical averages
The
Where
3-4
,
'
mill test report data exist for a grade of perimeter
combined with the NIST-measured
No
data.
column
correction for location effects or
was necessary.
static yield strength for rolled
literature data (Alpsten
shapes was calculated using the expected historical value from
1972) rather than the measured values, with correction for dynamic and location
effects:
<^.Vi
where
is
the specified
minimum
~ ^s^y + ^'dynamic + ^'nangc
yield strength.
historically, the mill test report value
The
factor
ks^
'
1.2 corrects for the fact that,
of the yield strength of rolled shapes exceeds the specified minimum
value by 20 percent. (Alpsten, 1972). The factor
Anange
= -2.6
ksi corrects for the fact that, until recently,
came from the higher-strength web, but the flange
The factor ^dynamic is calculated from Eq. 3-1.
the specimen for the mill test report
majority of the load-carrying area.
represents the
For core rolled plates no correction for location effects was necessary:
^ys
NISTNCSTAR
1-3D,
WTC
Investigation
~^p^y'^
^dynamic
'
65
Chapter 3
where
kp
=
1.092 (Alpsten 1972). Table 3-13 summarizes the estimated static yield strengths for
steels in the fire
The second
step in estimating constitutive behavior
There
yield.
is
no
for all
plastic regime.
the
Voce hardening law
In this
to estimate the
work hardening of the
on work hardening behavior of WTC
steel after
steels, so the
WTC steels was modeled using stress-strain data generated in the
Investigation. Early in the Investigation, based
selected the
was
historical or literature infonnation
work hardening behavior
NIST
all
and impact zones.
on requests from other Investigation
to represent the increase in stress, a,
hardening rule the flow
stress, c,
Team members,
with plastic
reaches a limiting stress,
,
strain, Sp, in the
with an
exponential decay:
3-7
This relation
is
available in most finite element packages.
^
Table 3-13. Estimated static yield strengths and work-hardening parameters, Eq. 3-5, for
f
Description
All
Fy =36
ksi perimeter
ksi
TS
ksi
ksi
Rn
b
''max
(\
OA
yu
column steels in
- plate 3)
ib
jj.D
D1
column steels in
- plate 3)
45
53.1
74.9
0.153
column steels in
- plate 3)
50
54.0
75.6
column steels in
- plate 3)
55
60.8
column
60
1
it\
111
ksi
14 Ao
ksi
1
D /.y /3
24.965
27.198
62.245
0.220
28.659
27.870
74.790
82.6
0.259
18.479
24.698
76.770
62.0
87.3
0.176
27.535
24.543
74.925
65
69.6
90.4
0.138
38.284
23.847
89.520
70
76.7
92.0
0.103
19.499
26.777
10.714
column steels in
- plate 3)
75
82.5
96.8
0.070
29.008
17.463
17.826
column
80
91.5
99.4
0.079
32.567
14.203
29.522
All Fy
=85 Fy =90 Fy =100 ksi perimeter
column steels, regardless of plate
85
104.8
116.0
0.081
13.857
32.500
1.793
Perimeter Plate 3 (inner web)
42
42.6
67.2
0.153
24.974
26.843
62.110
Perimeter Plate 3 (inner web)
45
45.9
69.8
0.153
24.977
26.785
62.147
Perimeter Plate 3 (inner web)
50
51.4
74.2
0.220
28.777
24.729
73.369
Perimeter Plate 3 (inner web)
55
56.9
78.5
0.259
18.455
25.244
76.857
plates
1
2 4
All Fy
-45
plates
1
All Fy
-50
plates
1
All Fy
=55
plates
1
(i.e..
ksi perimeter
2 4
(i.e..
(i.e..
(i.e..
with thickness
All Fy
=60
<
All Fy
=65
plates
1
2 4
(i.e..
<
0.5
in.
with
All
t
Fy =70
1
All F,
=75
plates
1
All Fy
=80
1.5 in.
steels in
not inner plate - plate 3)
ksi perimeter coluinn steels in
not inner plate
-
plate 3)
ksi perimeter coluinn steels in
2 4
plates
not inner plate
ksi perimeter
12 4 (i.e..
witht< 1.25 in.
plates
(i.e..
not inner plate
ksi perimeter
2 4
(i.e..
1
not inner plate
ksi perimeter
2 4
u.
not inner plate
ksi perimeter
2 4
.1
not inner plate
-
plate 3)
not inner plate
ksi perimeter
steels all
plates
66
NISTNCSTAR
1-3D,
WTC
Investigation
Room-Temperature Tensile Properties
F,
"is
TS
ksi
ksi
ksi
Perimeter Plate 3 (inner web)
60
62.4
82.9
Perimeter Plate 3 (inner web)
65
67.9
Perimeter Plate 3 (inner web)
70
Perimeter Plate 3 (inner web)
75
>
Description
a.
is
h
ksi
ksi
0.176
27.547
24.298
74.922
87.3
0.138
38.407
21.774
88.546
78.9
96.0
0.070
31.698
29.734
0.000
78.9
96.0
0.070
31.698
29.734
0.000
•'max
the true strain at the tensile strength, TS.
The procedure
for determining the parameters of Eq.
For each experimental
1.
3-7 was multi-step:
stress-strain curve for a
elongation or truncate curve
at the 0.2
given
steel,
remove
yield point and yield point
percent offset yield strength for stress-strain curves
without this yield point behavior.
at the tensile strength, TS,
which
and
strain.
2.
Truncate the curve
3.
Shift the stress-strain cur\'e to zero stress
4.
Fit the
5.
For the
which
is
the beginning of necking.
parameters of Eq. 3-7 to the resulting curve segment.
set
is
of replicated
tests,
determine the smallest of the strains
at tensile strength, e^^x,
the point at which necking begins.
6.
Create a fitted plasticity curve for each experimental curve that terminates
7.
Numerically average each family of fitted curves
at
at Cmax.
each strain point to produce an average
behavior.
Refit the parameters of Eq.
8.
3-7
to the
average curve produce the average work-hardening
properties.
Tables 3-13 and 3-14 and Figs. 3-24 to 3-27 summarize the representative tensile stress-strain curves for
the 33
WTC steels characterized in the Investigation.
elastic
modulus (205 GPa = 29.7 Msi) extended
which
this
occurs will be close
to,
elastic portion
to the value
of the
Bolts
The
column connections between
exterior wall panels
The
strain at
ASTM A 325 or A 490 standards. The joints
A 490 bolts, and those higher up of four A 325 bolts.
The
bolts
either four or six high-
were specified
to
conform
to
lower on the building were generally composed of
either the
structural
static yield strength.
were fastened using
strength bolts, depending on the vertical location on the building.
six
of those curves use the value of
but not equal to the 0.2 percent offset yield strength, YS.
3.5.2
vertical
The
assembly
in
buildings and other structures, and
These bolt types are designed
for use in general
are not designed for either high-temperature or
severe environmental exposure.
Bolts are characterized by a smooth section, the shank, and a threaded section. Because the two sections
have different diameters, the load bearing section
NISTNCSTAR
1-3D.
WTC
Investigation
is
not uniform along the length of the bolt.
As
a
67
Chapter 3
consequence, the tensile stresses that each section of the boh feels are not uniform, and are higher
in the
threaded area. In addition, the threads complicate the characterization of the mechanical behavior
in three
ways.
First, the roots
of the threads concentrate stresses in those areas and cause preferential locations for
both plastic flow and fracture
making
the mechanical joint
initiation.
is
Second, the interaction of the bolt threads and the nut threads in
complex. The
first
three or four tlireads toward the bolt shaft take
up most
of the load of the joint. Finally, the threads deform under shear, rather than under tension, Fig. 3-28.
These
peculiarities both require the
mechanical properties of bolts to be specified differently from
ordinary structural steel and introduce additional variations in bolt performance in service.
The use of stress-strain
capture their behavior.
constitutive behavior to describe the mechanical behavior of bolts does not fully
The aforementioned
mechanical behavior be described
that the
effects of the threads in localizing plastic response require
in
terms of load-elongation curves from a tensile
actual bolt and nut sample. Nevertheless, for each bolt diameter
tensile strength in
pounds derived from a minimum
tensile strength,
load-bearing area in the threads for bolts less than 1.125
rather than
in.
specifies a
TS = 120,000
diameter. For
test
of an
minimum
ksi, calculated
on the
ASTM A 490 bolts, the
minimum values for yield strength, F, tensile strength, TS, total elongation to failure
of area, ROA. Tests on A 490 bolts are conducted on specimens machined from the bolts
standard specifies
and reduction
ASTM A 325
,
on the boUs themselves.
Table 3-14. Estimated static yield strengths and work-hardening parameters, Eq. 3-5, for
core column and truss steels.
TS
Description
ksi
ksi
0.126
20.523
31.539
0.000
74.1
0.190
31.113
20.583
64.369
37.0
n/d
0.190
30.337
24.467
67.973
42
53.8
n/d
0.202
23.023
32.116
30.495
ksi
ksi
Truss rounds specified as
36
38.1
59.6
Truss
36
55.3
36
A 36
angles (regardless of ASTM
Ro
b
ksi
*'niax
specification) and all rounds specified as
A
242
Fy =36
ksi core
Core Group
1
WF shapes
&2
shapes
Core Group
3 shapes
42
49.0
n/d
0.202
23.023
32.116
30.495
Core Group
4&5
shapes
42
44.2
n/d
0.202
23.023
32.116
30.495
Core Group 1&2 shapes
45
53.8
n/d
0.202
23.023
32.116
30.495
Core Group
3 shapes
45
49.0
n/d
0.202
23.023
32.116
30.495
Core Group
4&5
shapes
45
47.8
n/d
0.202
23.023
32.116
30.495
&2
shapes
50
53.8
n/d
0.202
23.023
32.116
30.495
50
53.8
n/d
0.202
23.023
32.116
30.495
36
36.7
64.5
0.204
21.723
30.729
58.392
42
51.4
n/d
0.202
23.023
32.116
30.495
42
47.0
n/d
0.202
23.023
32.116
30.495
42
42.6
n/d
0.202
23.023
32.116
30.495
Core Group
Core Group
1
3 shapes
All 36 ksi core box
Core Plates with
t
<
column
0.75
Core Plates with 0.75
Core Plates with
Key:
a.
68
1
in.
.5 in.
steels
in.
<
<
t
t
<
<
1.5 in.
4.0
in.
n/d, not determined.
Smax
is
the true strain at the tensile strength, TS.
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
0
I
1
I
I
0.00
I
I
I
0.05
I
I
I
I
I
0.10
I
I
I
True
14:40:01
-
I
I
I
I
0.15
I
I
I
0.20
1
I
I
1
I
0.25
I
I
I
0.30
strain
Saturday August 21 2004
,
Figure 3-24. Representative tensile stress-strain behavior for perimeter column steels
from flanges, outer webs, and spandrels (plates 1, 2, and 4).
Figure 3-25. Representative tensile stress-strain behavior for perimeter column steels
from inner webs (plate 3).
NISTNCSTAR
1-3D,
WTC
Investigation
69
Chapter 3
Figure 3-26. Representative tensile stress-strain behavior for selected core column
steels.
Q
I
I
0.00
I
I
I
I
1
0.05
I
1
I
I
0.10
I
I
I
I
True
14:40 13
-
I
I
1
I
0.15
I
I
I
0.20
I
I
I
I
0.25
I
\
i
I
0.30
strain
Saturday August 21, 2004
Figure 3-27. Representative tensile stress-strain behavior for truss steels.
70
NISTNCSTAR
1-3D,
WTC
Investigation
Room-Temperature Tensile Properties
Figure 3-28. Representation of deformation that occurs in exposed bolt threads
and in bolt threads coupled with nut threads (right).
In simulating the response of the
strain
bohs
behavior measured for the base metal of the bolt, mesh the bolt with
then generate the load-displacement behavior up to failure. However,
simply
test bolts in
independent of the strain
The
rate.
specimens taken from the same
all
was produced by calculating the average
each curve by truncating
at
The load values of the four
Chapter
all
failure
A
325
bolts,
shows
4,
produced similar
four tests and
is
to
that the bolt strength
consequence, an
plotted in bold. This average
tests,
then modifying
or extending to this displacement (in the case of extension, using a spline
tests
were then averaged
to
presented load-elongation data for
A
fit).
produce the bold curve, which should be used
room
in
temperature.
Relatively few experimental studies of A 325-type bolts have been published.
test.
more accurate
results. In
displacement for the four
calculating the mechanical response of the bolted joints at
during the
of its non-uniformities, and
easier and
high-rate Kolsky bar tests and quasi-static compression tests on
bolt, described in
average load-displacement curve was calculated from
curv e
it is
stress-
tension and supply the actual load-displacement curves.
Figure 3-15, a plot of the four load-displacement curves for the
is
one would take
as a feature of the finite element model,
(left)
Rumpf and
325 bolts and measured the actual change
in length
Fisher (1963)
of the bolt
Other smdies (KJrby 1995, Li 2003, Sakumoto 1993, Salih 1992) reported proof loads or
failure loads while investigating temperature or other effects
on strength.
WTC towers, Table 3-1, are higher than measured in
previous studies and much larger than the A 325 specification. The average tensile strength of the WTC
The
A
tensile strengths
325 bolts
is
of A 325 bolts from the
(67,700±300)
Ibf,
or 22 percent greater than the specified
minimum. The
failure loads
(Rumpf 1963) range from 53,000 Ibf to 67,500 Ibf
with an average of 58,500 Ibf or 10 percent greater than the specified minimum. When Rumpf et al.
conducted these tests, the A 325 standard was based on TS = 1 15,000 Ibf, so the specified minimum
taken from ten batches of 7/8
tensile load
in.
bolts in the 1960s
was 53,200 Ibf The strength of the
NIST NCSTAR
1-3D,
WTC
Investigation
WTC bolts is comparable to the
1990s era 7/8
in.
A
325
71
Chapter 3
bolts characterized
by Salih
et al. (1992),
17 percent greater than the specified
whose mean
tensile strength
was (65,000± 400)
Bolt strengths that exceeded the requirements of A 325 were apparently quite
values for the
would under
WTC bolts are unexpected.
Use of the
or
common,
but the high
literature values for the bolt strengths in the
column connections, and might tend
predict the strength of the vertical perimeter
incorrectly predict a global collapse
Ibf,
minimum.
models
to
mode.
Studies in the literature have established the effect of the
number of exposed
both ultimate load and elongation of the bolt. Figures 3-29 and 3-30
show
threads above the nut on
the results
from the
literature
studies graphically. Because bolts have a non-uniform cross section, deformation tends to concentrate
within the exposed threads above the nut. Increasing the
over which
this
deformation occurs and increases the
has the opposite effect. In terms of critical load
number of exposed
total
threads increases the length
elongation to failure. Exposing fewer threads
threads act as stress concentrators, and
at failure, the
failure will occur at the location within the threads that has the greatest stress concentration. Increasing
the
number of exposed
threads increases the
number of stress
concentrators, and increases the chance of
loading a more severe stress concentrator, which produces failure
et al.
at
lower loads. The data of Rumpf
(1963) indicate that the linear strength decrease stops above nine exposed threads, however.
1
O
m
m
55
— ——
———
I
1
I
I
—— —
1
I
i
1
I
— ——
I
— ——
1
I
I
I
= (-1.05±0.15) kip/thread: Maximum Tensile load
= (-1.10±0.16) kip/thread: Load at Yield
O
g.
-6
50
CO
o
45
Source: Salih 1992
J
I
I
I
1
I
I
I
I
I
L
J
3
2
Number
of
I
1
I
I
I
I
I
I
I
I
I
L
5
Threads Exposed
Figure 3-29. Tensile strength change of
with exposed thread length.
72
I
4
A
325 bolts
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
0.34
1
— —— ———
I
1
I
I
I
^
I
I
I
I
I
I
I
"I
I
I
I
I
I
I
I
I
r
I
I
o
o
0.32
o
CL
LO 0.30
Gi
O
"cc
0.28
-4—'
c
CD
E
(D
O
0.00
0.05
O.lO
0 15
0.20
displacement,
0.25
0,30
in.
0.26
o
o
O
O
0.24
-
0.22
o
Source: Salih 1992
J
0.20
0
I
L
J
3
2
1
I
I
Number
of
I
I
I
I
I
I
I
I
4
L
5
6
Threads Exposed
Figure 3-30. Displacement near failure of A 325 bolts
as a function of number of threads exposed.
The
Investigation generally followed the procedures in
ASTM F 606 for testing threaded fasteners which
mandates leaving four threads exposed during the load-displacement
deviated from
F 606
in that
only two threads were exposed, which
exposed threads on failed bolts seen
era,
A
370 governed the
absence of other data,
test
it is
A
325
bolts,
is
same
the
elastic
displacement curve of the bolt
in Fig.
2,500,000
Ibf/in.
3-3
1
exposed threads. In the
from contemporary materials.
measured load-displacement curves are
Both Rumpf (1963) and Salih (1992) used an
compliances of the load frame, grips,
itself,
the load-displacement curve
to increase the slope
(Rumpf
six
The data of Fig. 3-15 represent the displacement of the crosshead of the
machine, which includes the
modified
an average value of the number of
and mandated
as literature values
linear portions of the experimentally
bolt.
The study
reasonable to assume that the statistical variation of the percent change in
significantly lower than those published in the literature.
extensometer on the
is
bolts.
WTC recovered components. During the WTC construction
requirements for
strength with thread exposure
The slopes of the
in the
of heavy head
test
etc.
from the
To
testing
reproduce the load-
test. Fig.
3-15 has been
of the linear portion to match that reported in the literature,
1963).
Using Bolt Data
Use
the average curve of the actual tests. Fig. 3-31, for the
Using
literature values,
NISTNCSTAR
which are lower, may
1-3D, WTCInvestigation
A 325 bolts from the World Trade Center.
result in inaccurate failure
mode
predictions.
73
Chapter 3
Use
literature values for the linear portion
shows
A
of the load-displacement curve for
325 bohs. Figure 3-3
this corrected data.
— —— ——— — —— — —— — — — — — ——— — ——— — — — — —
I
3x10''
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0.10
0.05
0.00
0.15
0.30
0.25
0.20
displacement,
in.
Figure 3-31. Data for load-displacement of A 325
bolts from Fig. 3-15 corrected for initial elastic slope
from
Assume no
Fig. 3-15,
Use
strain rate effects
on strength, because the Kolsky data
do not show a strong
the data of Figures
literature.
in
Chapter
4,
and the
tensile data,
effect.
3-29 and 3-30
strength and elongation at failure.
to estimate the effect
Assume no
of different exposed thread lengths on tensile
further strength reduction for
more than nine exposed
threads.
3.5.3
Welds
To
modeling the
assist in
structural perfonnance. Tables
3-2 and 3-15 summarize the as-measured and
as-designed weld data. Table 3-2 contains data generated as part of this study. Table 3-15 contains data
for the strength ranges listed in the construction
documents.
Table 3-15. Room-temperature weld metal properties as designed.
For Applications With
F,
< 60
ksi
60 ksLF,75
F,
a.
74
> 75
ksi
ksi
Weld Metal
Fy = 65
TS = 80
ksi,
Fy =77.5
Fy=
Properties'
ksi,
107.5 ksi,
ksi
TS = 85
r5=
Theoretical weld properties (not including the strengthening due
by base metal).
ksi
115 ksi
to dilution
NISTNCSTAR
1-3D,
WTC Investigation
Room-Temperature Tensile Properties
SUMMARY
3.6
The
Investigation
made
several findings regarding the properties of the recovered steel and the original
specifications and expectations.
The
yield and tensile strengths of the perimeter columns in 22 of 24 cases
are larger than the
minimum
by which they
short are consistent with expected values for the average strength and coefficient of
fall
requirements. The number of slightly under-strength plates and the amount
variation of yield strengths of plate steels from the WTC-construction era.
strength to specified yield strength, approximately
WTC-construction
era.
1
.
1
is
,
possible that these low values arose from
measured yield strength and yield point
damage
that
minimum F,. = 36
A 36.
of measured yield
ksi specified in
for the wide-flange
specimens
were high-strength, low-alloy (HSLA)
steels,
is
A
specified in
36.
It is
about 10 percent lower than
era,
however.
Many
of the steels in
even when they were specified
and so were much stronger than specified. Recovered bolts were stronger than the
literamre.
A
removed the yield point behavior. The average
expected value predicted from the literature of the WTC-construction
to
ratio
Unlike the perimeter columns, the measured yield strengths of 2 of 8 samples of
wide-flange shapes recovered from the core are less than
the floor trusses
The
also consistent with literature estimates from the
to
conform
minimum
325, and were much stronger than expected based on reports from the contemporaneous
The welds
tested
had strengths consistent with the requirements of the design documents.
3.7
REFERENCES
3.7.1
References Available from Publicly Available Sources
AISI. 1974. The Variation of Product Analysis and Tensile Properties Carbon Steel Plates and
Flange Shapes,
in
Wide
Contributions to the Metallurgy of Steel. American Iron and Steel Institute.
Washington. DC.
Alpsten G. A. 1972. Variations in Mechanical and Cross-Sectional Properties of Steel,
Design of Tall Buildings volume
Banovic
the
S.
W., C.
McCowan
lb.
American Society of Civil Engineers.
Banovic
S.
755-805.
NCSTAR
1-3E. National
W., Foecke T. 2005b. Federal Building and Fire Safety Investigation of the World Trade
Damage and Failure Modes of Structural Steel Components. NIST NCSTAR
National Institute of Standards and Technology. Gaithersburg,
J.
New York.
of Standards and Technology. Gaithersburg, Md. April.
Center Disaster:
Barsom,
Planning and
and W. E. Luecke. 2005. Federal Building and Fire Safety Investigation of
World Trade Center Disaster: Physical Properties of Steels. NIST
Institute
in
M.
Md.
1-3C.
April.
1988. Characteristics of Heavyweight Wide-Flange Structural Shapes. Welding Research
Council Bulletin. 332. 1-19 April.
Barsom,
J.
M. 1987. Material Considerations
in Structural Steel
Design.
AISC Engineering Journal
.
24[3]
127-139.
Beedle, L.
S.
and L.
Elliot, R. A., E.
after aging,
NIST NCSTAR
Tall. 1960.
Orowan,
Basic
Column
Strength.
ASCEJ.
T. Udoguchi, and A. S. Argon. 2004.
Struct. Div.
WTC
Investigation
7.
139-173.
Absence of yield points on
and the Bauschinger overshoot. Mechanics of Materials. 36.
1-3D,
ST
1
strain reversal
143-1 153.
75
Chapter 3
Johnson, R. F. 1967. The Measurement of Yield Stress. In Strong tough structural steels: proceedings of
the joint conference organized by the British Iron
and Steel Research Association and the Iron and
Royal Hotel Scarborough, 4-6 April 1967.
Steel Institute at the
Kirby, B. R. 1995. "The Behaviour of High-strength Grade 8.8 Bolts in Fire."
Research. 33. 3-38.
Chen and M.-F.
Yin, K.
E, T. A. Siewert,
and
2003 "Experimental Studies on the Properties of
Li.
Constructional Steel at Elevated Temperatures."
J. Struct.
Eng. 129[12]. 1717-1721.
W Gayle. 2005. Federal Building and Fire Safety Investigation of the
F.
World Trade Center Disaster: Contemporaneous Structural Steel
Rao, N. R. N, M. Lohrmann, and L.
Rumpf,
J.
Journal of Materials.
and
L.
J.
W.
Specifications.
NIST
NCSTAR
of Standards and Technology. Gaithersburg, Md. April.
1-3 A. National Institute
Steels.
Construct. Steel.
J.
-
Li, G.-Q., S.-C. Jian, Y.-Z.
W.
Steel
London. 51-60.
Institute.
Luecke,
&
ISl Publication 104. Iron
1
Tall. 1966. Effect
of Strain Rate on the Yield Stress of Structural
24 1-26 1
[2]
Fisher. 1963. "Calibration of A325 Bolts."
J. Struct.
Div.
ASCE. 89[ST6].
215-234.
Sakumoto, Y., K. Keira, F. Furuyama and T. Ave. 1993. "Tests of Fire-Resistant Bolts and Joints."
J. Struct.
Eng.
1 1
9[
1 1 ].
3
1
3 1-3
1
50.
M. Aktan and M. Usmen. (1992) "An Experimental Appraisal of the LoadDeformation Properties of A325 High-Strength Bohs." J. Test. Eval. 20[6]. 440-448.
Salih, N.,
Smith, H.
J.
Tipper, C. F. 1952. Effect of Direction of Rolling, Direction of Straining, and Ageing on the Mechanical
Properties of Mild-Steel Plate.
Uko, D. R. Sowerby, and
J.
1.
Iron Steel
Inst.
172[10]. 143-148.
D. Embury. 1980. "Bauschinger Effect in Structural Steels and Role in
Fabrication of Line Pipe: Part
Technology.
J.
1
Analysis of Bauschinger Effect
in Structural Steels."
Metals
359-367.
Voce, E. 1948. "The Relationship between Stress and Strain for Homegeneous Deformation.
J. Inst.
Metals. 74 537-562.
WEL-TEN.
1966. Iron
WEL-TEN
steels.
Age 198 p
"Nickel
18 Advertisement
may
be added to
Woldman's Engineering Alloys. 1990. Frick
J.P. editor.
60 and 62
if
necessary." October
6.
ASM International. Materials Park, OH.
References Available from Nonpublic Sources
3.7.2
Betts, Jr,
showing representative properties of Yawata
WEL-TEN
W.
August
E. 1967. Letter to
25.
"We
R.M. Monti (PONYA)
that details
have had five representatives of Yawata Iron
several days discussing the Contract
Montague-Betts
&
Steel, Ltd.
which we are placing with them
for a
steel purchase.
of Japan
in
minimum of
our office for
12,000 tons of
wide flange beams and columns."
76
NIST NCSTAR
1-3D,
WTC
Investigation
Room-Temperature Tensile Properties
Document QC-101 "Quality Control Data Resistance Weld Designation and
Laclede. 1968.
March
23.
Laclede. 1968b
Memorandum No.
4. Interoffice
and drawings for bulb angles used
Laclede. 1968c
Memorandum No.
specifications
Laclede. 1969.
memorandum from R.D. Bay
to construct floor trusses.
and drawings round bars used
Memorandum from E.M.
certification.
December
Marubeni-Iida. 1967.
18.
January
memorandum from
11. Interoffice
that contains specifications
16.
R.D. Bay that contains
to construct floor trusses.
February 28.
Elgan. General Welding Foreman, to Fred Deppe, St Louis
Testing Laboratories, that contains qualification data on
fillet
welds for welder and procedure
NIST Document WTCI-509-LERA
Memorandum from
H.
Yamada
(Marubeni-Iida (America)) to
that contains copies of shipping orders for first three shipments
of steel plate
I.
as
A
Mitsui. 1968. Shipping invoice from Mitsui
&
Co. to Pacific Car and Foundry for 382.892
lists
conforming
steel mill.
June
Monti, R.M. 1967.
to
three plates
A 36-66 for spandrels.
572 Grade 42.
(PONYA)
Soffer
to Stanray Pacific.
marked
shipping order
plate
Strength."
The
5 pages, July 18.
Originating port was Chiba, Japan,
site
lb
of steel
of a Kawasaki
10.
Memorandum from R.M. Monti (PONYA) to William Clarkson (Montague-Betts)
Yawata "A 441 modified" steel composition. Includes supporting
that
contains approval to use the
correspondence from Yawata Steel to Montague Betts (dated November
2,
1967) with chemical and
mechanical requirements. 5 pages. November 21.
Morris, R. E. 1969. Letter to James White
plate
(SHCR)
from Fuji Steel Hirohata works. Aug.
PCF. 1967.
Pacific
that contains a mill test report
H9-5386
for an
A
36
5.
Car and Foundry. Documents WS-1 lA through WS-1 IC "Weld Procedure." June
1.
NIST document number WTCI-158-LERA
Skilling-Helle-Christiansen-Robertson. 1968. Inter-office
Stanray Pacific. 1967.
(PONYA)
May 23.
Memorandum from
that describes the use
Stanray Pacific. 1967b.
J.
White (SHCR)
April
VP
25.
of Engineering (Stanray
Pacific), to L. Feld
of tandem submerged arc welding to join core column plates.
Memorandum from
that describes the use
Stanray Pacific. 1968.
H. Kjerulf,
memo, October
of a self-shielded
Memorandum from
F.
Allen (Controller, Stanray Pacific) to L. Feld
(PONYA)
FCAW electrode for welding the ends of columns. June 2.
R.E. Morris (Contract Administrator, Stanray Pacific) to
that describes modifications to the
welding procedure SA-4 during the contract.
8.
Stanray Pacific, undated. "Stanray Pacific Welding Procedure Qualification SA-2 (revised)," Report
includes transverse
NISTNCSTAR
fillet
weld strength
1-3D, WTCInvestigation
data. Filed
with documents dated October 1967.
77
Chapter 3
PCF Memo #666-90 Control Procedure for Material Substitution. Memorandum from
R.C. Symes (PCF) to R.M. Monti (PONYA) that details the procedure for substituting steel with yield
strength larger than specified. May 29.
Symes. R.C. 1968.
Symes R.C. 1969 Memorandum PCF D-666 #
Construction, FebiTiary
5.
M. Gerstman, Tishman Realty and
was that plates #1, 2 and 4
#3 would be domestic. In fact, we have
T-40. Letter to
''Our original buying plan for this contract
would be imported and stripped
our plant, while plate
in
occasionally bought minor quantities of domestic material instead of planned imported, to maintain
we have adhered
fabrication schedules and so on, but in general
Symes. R.C. 1969b.
(PONYA)
PCF Memorandum D-666
up
to the present
in
our original concept."
Memorandum from Symes (PCF)
-#T-178.
to
R.M. Monti
of a Yawata Iron and Steel investigation (dated Dec. 18, 1968) into
that contains the results
"minor laminar discontinuities"
to
A
36 spandrel
plates, "all
FY
36 spandrels incorporated
into panels
time meet the contract specifications for this material and that discontinuities found
do not represent injurious defects."
Warner, H. L. 1967. Letter
July
7.
The
memo
to
Malcolm Levy (PONYA)
shows 10,240 tons of "British"
that details Stanray Pacific steel
steel 1.75 in.
purchasing plan.
and thicker and 21,760 tons of
"Japanese" steel of all gauges. Other documents identify Fuji as the Japanese supplier and Colvilles
as the British supplier.
White,
J.
1967.
for Pacific
Memorandum from James White (SHCR)
Car and Foundry. September
Kawasaki and Yawata
White,
J.
1967b.
6.
to
"Infomiation
are producing steel for
Memorandum from James White (SHCR)
White,
Fy
78
J.
WEL-TEN
60,
1968. Letter on
=100
WEL-TEN
62,
WEL-TEN
SHCR letterhead to R.M.
ksi steel for F,.
= 90
now
in
our hands indicates that only
PCF."
to
changes to original specification for proprietary Yawata
60R,
R.M. Monti (PONYA) on sources of steel
70,
R.M. Monti (PONYA)
steel grades,
WEL-TEN 80C
Monti
(PONYA)
that details required
"A441 -modified,",
(described as
"A
WEL-TEN
514-modified.")
that approves the substitution
of
ksi in all instances. Feb. 15.
NISTNCSTAR
1-3D,
WTC Investigation
Chapter 4
High-Strain-Rate Properties
introduction
4.1
It is
well
known
that the yield
and ultimate strengths of steels increase with increasing
strain rate
(Harding 1979. Hutchinson 1977, Wiesner 1999). In terms of the mechanical response of the steels in the
towers, the high-speed impact of the aircraft
is
which
a high-strain-rate event,
is
well above the
conventional strain rates normally associated with mechanical properties measurements.
that the strain rates
1000
s"'.
on the World Trade Center (WTC)
Accurate models of the
aircraft
of the mechanical properties of the
The goal of the
steels,
due
to the aircraft impacts,
It is
estimated
were up
to
impact and the resulting damage to the towers require knowledge
steels at
high-strain-rate testing plan
high strain
was
to
rates.
develop a database of high-strain-rate properties on
WTC steels for use in modeling the extent of structural damage to the perimeter and core columns as a
result
of the aircraft impacts. For strain rates up
on eight perimeter- and
five
core-column
sensitivity for the Investigation.
tests,
using a
split
steels.
For higher
rates
to
500
s'',
These data were used
up
to
Hopkinson pressure bar or Kolsky
4.2
TEST PROCEDURES
4.2.1
High Strain-Rate Tension Tests
2500
test,
employed high-rate tension
the test plan
s
'
to
the plan
tests
develop models for strain rate
employed high-rate compression
on one bolt and two perimeter-column
steels.
High-strain-rate tension tests were conducted on eight perimeter-column steels and five core-column
steels at rates
of
1
.0 in., a
between 50
width of 0.25
s"'
and 500
in.,
s
'.
High-strain-rate tension specimens. Fig. 4-1,
and a thickness of either 0.125
in.
or 0.063
in.
had
a
gauge length
was measured using
Strain
open-grid, high-elongation strain gauges attached to the specimen test section with high-elongation strain
gauge adhesive.
An
additional strain gauge attached to the fixed-grip end of the specimen
measure the load and
to validate that the
specimen grip-end did not deform
was used
to
plastically.
0.125
7.00
0.053
TK
Uniform Test Section
Flat-HSR-1 "
TK
Dimensions are
Flat-HSR-1 "
in
TN
Inches
TN
Figure 4-1. Specimen used for high-strain-rate tension tests.
NISTNCSTAR
1-3D,
WTC
Investigation
79
Chapter 4
The
machine
high-strain-rate test
actuator
mounted
in a
1
is
a closed-loop, servo-hydraulic
10,000 lb load frame (Bruce 2003).
machine with an
11, 240 lb capacity
A 400 gal per min. servo-valve, suppUed by a
A digital controller and associated
5 gal capacity accumulator, controls the high rates for this machine.
computer system provide system
control.
load and 10 m/s at 50 percent of the
The
specified peak velocity for the system
maximum
is
13.5
m/s
at
zero
Data are acquired using commercially available
load.
software and a data acquisition board capable of acquiring 500,000 data points per second. The typical
acquisition rate
is
10
ms
per point, which allows 5,000 data points per
measurement devices: a piezoelectric load washer attached
The system employs two load
test.
to the fixed
end of the load
train
and a
strain
gauge bonded to the grip end of the specimen.
A
slack-adaptor in the load train. Fig. 4-2, allows the actuator to achieve the desired displacement rate
before loading the specimen. The tension specimen
is
gripped using bolt-tightened
wedge
grips.
Triggering the system starts the data acquisition and accelerates the actuator to the desired testing speed.
At the end of its
travel, the slack adapter
engages the lower grip rod and loads the specimen
at the
desired
extension rate. The targeted strain rates were not always achieved, but the resulting rates could be
calculated and are used in this report.
Piezoelectric load
'
1""^
'"
^
^
'
i
I
Slack adaptor
Figure 4-2. Schematic diagram of the slack adaptor apparatus.
Data collected during high-strain-rate tension
tests include the load
the actuator displacement, the strains in the test section
gauges, and the time.
strain vs.
80
The reported
strain rate realized
time curve recorded for each
from the piezoelectric load washer,
and the grip-end of the specimen from the
by
the specimen
is
strain
obtained from the slope of the
test.
NISTNCSTAR
1-3D,
WTC Investigation
High-Strain-Rate Properties
Analysis of High-Strain-Rate Tension Test Data
4.2.2
The data were analyzed using
1
.
a multi-step technique (Bruce 2003):
Correct the specimen gauge section and grip section strain data for Wheatstone bridge nonlinearity.
2.
Plot the engineering strain,
e,
Determine the engineering
strain rate,
linear part
time increment, At
,
for
ti
Convert the grip-section
t\
in the plastic region
<i <
by
fitting a line
=
e(t)
strain data to load,
+
to align
washer
4.
with the
it
at
each
At-e^,
4-1
using the calculated modulus and cross-sectional
area of the grip-end of the specimen. Overlay the grip section load curve
piezoelectric load
through the
Calculate the engineering strain
as
e(t-\-At)
3.
.
of the data over a time range
/ >
using the gauge-section strain gauge data.
versus time,
cur\'e, and, if
mean of the
on the data from the
necessary, adjust the modulus in the grip load signal
load washer signal to produce the corrected load signal.
Calculate the engineering stress based on the corrected load signal and the cross-sectional
area of the test section of the specimen.
5.
Calculate the true strain and true stress values from the engineering strain and engineering
stress values.
6.
Choose the uniform
7.
Compute
strain limit as the strain value at the tensile strength.
the total elongation, £/„ or engineering strain at fracture,
between gauge marks made on the specimen prior
El,
specimen
test section prior to testing,
and
ROA =
9.
following testing,
Lo,
Lf.
4-2
ROA, from
the reduction of area,
from the distance,
'
^
Compute
to testing to that
e/,
the original cross-sectional area, Ao, of the
measured
that
An ~ A
—
after testing,
Ap
f
4-3
Plot the engineering- and true-stress-strain curves
up
to the point of instability (uniform strain
limit).
10.
Determine the yield strength by visually estimating the mean value of stress just prior
to the
onset of strain hardening. In cases where the ringing of the load signal precludes reliable
visual estimation,
( 1
NISTNCSTAR
percent
1-3D,
<e<5
WTC
fit
a third order
polynomial to the yielding and
percent) of the curve.
Investigation
Draw
strain
hardening regions
a line parallel to the linear-elastic part of the
81
Chapter 4
curve
stress-strain
an offset of
at
1
The
percent.
intersection of these
the estimated yield strength, see Fig. 4-3. The value of
maximum
downside of the peak
significantly
skewed the
data.
load curve often
Dynamic
for
lines is
percent was chosen because the
1
fell slightly less
The ESIS Procedure
two constructed
than this offset point and
Tensile Tests (2000)
describes the procedure in greater detail.
1 1
.
Determine the tensile strength, TS, by dividing the
maximum
load by the original cross-
sectional area of the test section.
150
w
100
3rd order
polynomial fit
for
0.01<e<0.05
50
1
I
I
0.00
I
I
I
0.02
\
I
I
I
True
10.57,25
-
Friday February
1 1
,
I
I
I
I
I
0.06
0.04
I
I
L
I
0.08
0.10
strain
2005
Figure 4-3. Schematic of the procedure for
estimating the tensile yield strength when ringing in
the load signal precludes reliable visual estimation.
Kolsky Bar Tests
4.2.3
Tests at very high strain rates employed the split Hopkinson pressure bar, or Kolsky,
test.
method
is
example Follansbee
(1985).
The device
well known, and the literature of high-rate testing describes
HRC) maraging
striker bar at the
as
shown
in detail, for
consists of two hardened steel bars lightly supported in a frame as
schematically in Fig. 4-4. The test specimen, which
hardened (58
it
steel bars, 15
end of the input
in Fig. 4-5. This pulse,
bar.
is
lubricated to reduce friction,
mm in diameter and
The impact generates an
which
travels
strain
down
placed between two
compressive pulse
in the input
bar
the input bar at nearly 5000 m/s, compresses the test
gage
Input bar
shown
m long. A gas gun fires a 0.5 m long
elastic
strain
m
striker bar
1.5
is
This test
gage
m
specimen
Output bar
Figure 4-4. Schematic diagram of Kolsky bar apparatus.
82
NISTNCSTAR
1-3D,
WTC
Investigation
High-Strain-Rate Properties
I
I
I
I
I
0.010
Input bar
Output bar
11
PV\A
CO
~
0.005
Reflected strain
wave
in
(proportional to strain in
Input bar
^
specimen
|
CD
0.000 -
e
0)
re
o
>
-0.005
Transmitted
st.'ain
wave
(proportional to stress
-0.010
oading stress wave
2.4
2.6
2.5
2.7
2.8
t,
J:16
46
-
Tuesday December
14,
in
in
in
output bar
specimen
input bar
2.9
3.0
3.1
ms
2004
Figure 4-5. Oscillograph record of an incident pulse
that is partially transmitted to the output bar and
partially reflected in the input or incident bar.
specimen between the input and output bars
at strain rates
between 200 s"'and 10,000
s"'.
Thin copper
disks placed between the striker bar and the input bar shape the loading pulse and reduce oscillations
during dynamic loading of the specimen. The exact strain rate depends on parameters such as the length
and velocity of the
striker bar,
and the
specimen dimensions. Some of the loading pulse on the
test
specimen enters the output bar as a transmitted
see Fig. 4-5.
bars,
The
reflected
and transmitted
strain
strain
wave, and some
waves
dynamic Wheatstone resistance bridges, and dual
specimen
at
time
/
is
is
reflected
back
into the input bar,
are recorded using strain gauges
trace oscilloscopes.
related to the transmitted strain, er, in the output bar
The
test
mounted on the
stress, a, in the test
by
4-4
A
where
E is
Young's modulus of the
sectional area of the test specimen.
reflected strain,
£/?,
in the incident
bar,
The
Ao
is
the cross-sectional area of the bar, and
strain rate in the test
is
C is
the longitudinal
at
any
A
is
sound speed
in the
^
-2
C
—
£Jt)
bar and L
,
is
integrated to obtain the strain at any time during the test.
,
4-5
the length of the test specimen.
Combining
range 0.01 < £ < 0.20 are usually considered reliable. At strains below
and early yielding regions, the
than this range, the test specimen
NISTNCSTAR
1-3D.
test
specimen
may deform
WTC Investigation
stress has not
The
strain rate
the strain rate data with the stress
record provides the high rate constitutive behavior, or stress-strain curve. In the Kolsky
elastic
the cross-
instant is related to the
bar by
d£
—
=
where
specimen
this range,
test, strains in
the
which includes the
reached equilibrium. At strains greater
nonuniformly.
83
Chapter 4
The Kolsky bar
test
longitudinal axis
specimens were cylinders, 4
was
A
325
bolt.
line
column
line 141
summarizes the
between
results
floors 97-100.
of the bolt
plate.
Specimen C22-C2M-FL3
157 between floors 93-96. Specimen
column
of the
parallel to the rolling direction
perimeter columns and an
The
N8-C3M-FR2
Specimens originated from two
is
is
= 75 ksi plate from WTC
a F, = 60 ksi plate from WTC
a F,
original location of the
A
325 bolt
is
Because the lowest
1
unknown. Chapter
3
tests.
strain rate that the
Kolsky
test
can achieve
is
about 200
s
',
quasi-static
using Kolsky specimens were conducted to extend the strain rate data to low
employed an ordinary servo-hydraulic
most
1
Quasi-Static Compression Tests
4.2.4
tests
mm or 6 mm in diameter and 2 mm high, whose
cases, the
specimen
strain
test
machine
was estimated from
machine compliance. Tests on the
A
325 boh
steel
in
extension control
at
rates.
compression
These
tests
constant extension rate. In
the crosshead displacement, after correcting for
employed an extensometer clipped
to the
ends of the
loading rams as well.
4.3
RESULTS
4.3.1
High Strain-Rate Tension Tests
Figure 4-6
is
an example of a
M26-C1-RF, from
stress
WTC
1
set
of high-strain-rate stress-strain curves for
column 131
measurement was taken from
the stress signal
is
floors 90-93, north face, just
the strain
steels
and
column
gauge on the grip-end of the
test
specimen, the ringing in
not completely eliminated.
five
stress-strain curves for
84
ksi perimeter
above impact zone. Although the
Tables 4-1 and 4-2 summarize the results of the high strain-rate tension
column
= 50
F,.
core-column
each
steels.
Appendix
A tabulates the
tests
of the eight perimeter-
individual test data and presents the
test.
NISTNCSTAR
1-3D,
WTC
Investigation
High-Strain-Rate Properties
Table 4-1.
Summary
specimens and results
perimeter columns.
of
for high strain-rate tests of
Location^ in
Building and
Specimen
m
Shape
Orientation
Eq. 4-9
Specimen
(ksi)
M26-C1-RF
50
9/16
F,:
w=0.0121±0.0033
TS:
w=0.0118±0.0023
F,:
w=0.0178±0.0044
longitudinal
WTC
131 90-93
1
in.
flange plate
transverse
TS: W7=0.0 15810.0026
N8-C1-RF
60
5/16
F,:
w=0.0160±00033
TS:
w=0.0133±0.0032
F,:
w=0.0115±0.0032
w=0.0 12710,0012
longitudinal
WTC
143 97-100
1
in.
flange plate
transverse
TS:
F,:
N99-C3-RF
65
longitudinal
WTC
147 99-102
1
TS:
1/4 in. flange plate
w=0.0125±0.0023
w=0.01 3 1+0.0028
F,.:
w=0.0094±0.0036
TS:
w=0.0112±0.0022
transverse
F,:
A«=0.0067±0.0021
TS:
w=0.0138±0.0003
longitudinal
C22-FL
75
WTC
1
1
1
57 93-96
/4 in. flange plate
F,.:
OT=0.0061±0.0016
transverse
TS: w=0.0145±0.0038
Fy.
C14-SP
75
WTC 2
j/6
TS: »7=0.0 10810.00 15
300 86
spandrel plate
in.
w=n/d
longitudinal
F,:
transverse
TS:
w=0.0083±0.0021
w=0.0098±0.0016
F,:
w=0.0053±0.0014
TS:
w=0.0085±0.0013
longitudinal
MIOB-RF
100
14
flange plate
in.
F,:
w=0.0039±0.0038
transverse
TS:
ClO-IW
100
F,
WTC
1
451 85-88
:
/4 in. inner
web
w=0.0077±0.0033
longitudinal
TS:
1
w=0.0083±0.0010
w=0.0113±0.0006
plate
transverse
n/d
F,:
w=0.0044±0.0014
longitudinal
CIO-FL
100
WTC
1
451 85-88
1/4 in/ flange plate
TS:
w=0.0080±0.0026
F,
w=0.0051±0.0014
:
transverse
TS:
a.
WTC XXX
Key:
ft
YY-ZZ:
ff=tower
1
or 2:
XXX=column
n/d, not determined, generally because
line;
w=0.0091 ±0.0021
YY-ZZ=floor range
of a lack of low strain rate data
Note: Reported uncertainties are defined as the estimated standard uncertainty of w,
NISTNCSTAR
1-3D.
WTC Investigation
s,
85
Chapter 4
Summary
Table 4-2.
of specimens and results for high strain-rate tests of
core CO umns.
Location" in
ksi
36
Building and
Specimen
m
Shape
Orientation
(Eq. 4-9)
Specimen
C65-WEB
WTC
1
F,:
OT=0.0336±0.0030
TS:
w=0.0207±0.0010
longitudinal
904 86-89
web of 12WF161
transverse
36
F,:
WTC
C65-FL
n/d
1
flange of
w=0.0290±0.0042
longitudinal
904 86-89
TS: OT=0.0130±0.0014
12WF161
transverse
n/d
F,:
36
B6152-2-FL
1
TS: /w=0.0165±0.0012
504 33-36
flange of box column
F,:
transverse
36
1
«;=0.0317±0.0027
TS: ff?=0.0175±0.0039
14WF184
transverse
42
HHl-WEB
n/d
F,.:
WTC
1
w=0.0265±0.0014
longitudinal
605 89-101
TS: w=0.0200±0.0058
web of 12WF92
transverse
a.
WTC # XXX
Key:
YY-ZZ: #=tower
1
n/d, not determined, generally
or 2;
XXX=column
line;
because of a lack of low
n/d
YY-ZZ=floor range.
strain rate data
Note: Reported uncertainties are defined as the estimated standard uncertainty of m,
Figure 4-7
is
an example of the results of a Kolsky
and the corresponding
not constant during the
86
s.
High Strain-Rate Kolsky Bar Tests
4.3.2
steels
96+0 001
longitudinal
603 92-95
flange of
w?=0.0396±0.0009
rc- rii=0 01
F,:
WTC
C80
»;=0.0239±0.0038
longitudinal
WTC
and the
test.
The
plot
strain-rate-strain behavior for a perimeter
test.
A 325 boh
shows both the
column
Table 4-3 summarizes the Kolsky bar
steel.
Appendix
steel.
tests for the
A presents the results of the
other
stress-strain
Note
behavior
that the strain rate
is
two perimeter column
tests.
NISTNCSTAR
1-3D,
WTC
Investigation
High-Strain-Rate Properties
120
100
80
(0
0)
60
(1)
(2)
(3)
c
40
(4)
'cn
c
(5)
L8-06 de/dt=417 1/s
L7-06 d£/dt= 260 1/s
L6-06 dt7dt= 100 1/s
L2
di:-.'dt= 65 1/s
LI
d£ydt= 6.06x1 0"M/s
(4)
11(2)
(3)
(1)
LU
20
M26-C1B1-RF
WTCl Column
-I
I
130 90-93
L.
0.0
0.1
0.2
0.3
0.4
Engineering strain
10:1450
-
Friday February ll.
2005
Figure 4-6. Examples of tensile high-rate stressstrain curves for Fy = 50 ksi perimeter column steel
M26-C1B1-RF.
Figure 4-7. Example stress-strain and strain rate-strain
curves for Kolsky tests.
NISTNCSTAR
1-3D,
WTC
Investigation
87
Chapter 4
Table 4-3.
Summary
Test Specimen
Nominal Fy
and Location
(ksi)
of Kolsky bar tests.
Average
Flow
Average
Stress"
Rate''
Test #
(MPa)
(1/s)
434
943±43
15621408
435
895±55
11341123
436
804±46
579142
437
675±65
389195
438
1015+43 21021283
439
1029±31 25711423
440
829+42
441
835±53
1
450
831164
15331159
451
808164
1609180
452
776171
1016181
453
621182
7591119
454
453195
443187
455
860152
21691204
456
906155
23541246
457
921155
27991344
458
898156
26891206
324
691136
9171335
/IOC
C22-C2M-FL3
WTC
1
Strain
100
93-96
N8-L3JVl-rR2
WTC
529176
1411125
1
Column
60
141
97-100
1Tc
A 325 bolt
unknown
325
7951179
71
326
8361227
6881445
327
992170
11451338
328
957186
10601382
92
location
329
a.
Average flow
b.
Average
0.2en,ax
<
11434
stress represents the
strain rate represents the
e
<
0.8
1077144 23951566
average
average
in the
in the
range
range
Eniax-
Note: Uncertainties represent the standard deviation of the value;
\N-]
4.3.3
'
Quasi-Static Compression Tests
Table 4-4 and Fig. 4-8 summarize the quasi-static compression
for analyzing the strain rate sensitivity
1
0 percent
88
from the Kolsky bar
test
tests.
conditions and results that are used
The
plots are graphically truncated at
strain.
NISTNCSTAR
1-3D,
WTC Investigation
—
—
High-Strain-Rate Properties
Summary
Table 4-4.
of quasi-static compression tests
Initial Strain
Specimen
(ksi)
60
Location in Building
WTC
N8-C3M-FR2-1
5/16
142 97-100
1
in.
Rate
(1/s)
0.001
flange plate
0.01
75
WTC
C22-C2M-FL3
1
157 93-96
0.01
1/4 in. flange plate
100
WTC
C10-C1M-FL2
1/4
n/a
A
— ^
I—
0.01
0.002
XXX, column
or 2;
1
1—
line;
YY-ZZ,
r—
1—
I
1
\
1
1
(2)
N8-C3M-FR2-2
N8-C3M-FR2-1
(3)
A 325
(4)
A
(1)
^
451 85-88
flange plate
unknown
325 bolt
Key: WTC n XXX YY-ZZ: #=Tovver
n a = not applicable.
1500
1
in.
0.001
1
1
1
1
diVdt = 0.01
I
floor range;
I
1
1
1
s"'
ds/dt = 0.001
s"'
from machine compliance de/dt = 0.0023
325 bolt strain from extensometer
bolt strain
s"'
1000
c/5
3
500
0.00
1500
-|
0.04
0.02
1
r
True
0.06
strain
0.08
0.10
—
I
-I
1
T"
de/dt = 0.01 s
(2) C22-C2t</1-FL3-1 de/dt = 0.01 s
(3) C22-C2M-FL3-2 dtVdt = 0.001 s'
(1)
C10-C1M-FL2-1
'
^
1000
CO
3
500
0.00
0.02
0.04
True
0.06
0.08
0.10
strain
Figure 4-8. Quasi-static compression stress-strain
curves for the tests summarized in Table 4-4.
NISTNCSTAR
1-3D.
WTC
Investigation
89
Chapter 4
DISCUSSION
4.4
There are several important limitations
testing rates greater than about
200
s"'
in
understanding the data from the high-rate tensile
measurement of the
the specimen through the loading train begins to influence the
rates, the
time necessary for the
test
specimen
to
low
test.
stress
wave propagates through
test load.
At these
In this case, the experimentally derived
behavior does not represent the defonnation behavior of the
strain
At
achieve an equilibrium stress state throughout the gauge
section can be a significant portion of the total time of the
which the
tests.
(Follansbee 1985b) the propagation of the stress wave that loads
steel.
the specimen (Follansbee 1985b)
An
is
upper
limit for the rate at
the elastic
wave
velocity
v^:
4-6
where
E is
the elastic
modulus and
p is
A
the density.
lower limit
y
where dcr/ds
the stress
strain
is
wave
ds
the velocity of the plastic stress wave:
is
4-7
y
the instantaneous slope of the true-stress tme-strain curve. Follansbee (1985b) asserts that
velocity
is
closer to the elastic
wave
velocity even after yielding.
The minimum
can be estimated (Follansbee 1985b) from the time necessary for three passages of the
valid
wave:
stress
''min
through the specimen of gauge length lo deforming
1
percent offset strain used for the yield strength
stress state in the
is
at a strain rate
more than
£ For the high
.
rate tension tests, the
sufficient strain to ensure an equilibrium
gauge section.
A second issue in high-rate tests
is
the heating of the test specimen
by
work of plastic deformation
the
(Follansbee 1985b). Quasistatic tests are often assumed to be isothermal, where the heat generated by the
work of deformation can escape from
assumed
to
the specimen via the grips. In contrast, high-rate tests are generally
be adiabatic. In the short time of the
escape the specimen.
An upper
under adiabatic deformation
is
test,
the heat generated in defonning the specimen carmot
limit estimate, following Follansbee (1985b), for the temperature rise
about 80 °C. The Investigation
made no
attempt to correct for changes in
yield and tensile strength or elongation resulting from adiabatic heating.
Ringing of the load
The
signal. (Gillis
1985) introduces a third problem in interpreting the stress-strain curves.
ringing generally occurs at about the natural frequency of the load
particularly severe for high-rate tests that
employ conventional load
gauges
cell.
cells
away from
The problem can be
(Bruce 2004). Even with
test
specimens instrumented with
strain
used
not always possible to eliminate the ringing (Bleck 2000; ESIS 2000;
in the Investigation,
it is
Wiesner 1999; Yan 2002). Many
90
tests
in the grip section
the load application, like those
reported in the literature have been heavily filtered to remove the
NISTNCSTAR
1-3D,
WTC Investigation
High-Strain-Rate Properties
load ringing (Gillis 1985). In reporting the data, the Investigation did not attempt to mathematically
remoN
e the ringing.
Calculation of Strain-Rate Sensitivity for Tension Tests
4.4.1
Strain-rate sensitivity
(SRS)
is
an indication of the effects of strain rate on material properties such as
yield and tensile strength. This report defines the
Although there are other representations of the
effect
of strain
rate
on
yield, F,
.
and
SRS
as the slope of the log(stress)-log( strain rate) curve.
stress-strain rate behavior, this analysis expresses the
tensile strength, TS, via a simple relation:
^8^
4-9
(7 (I
where a
is
the yield, F,
V ^0 J
or tensile, TS, strength, ao
.
reference strain rate, and
m
is
=
1
ksi is a normalizing parameter,
the strain rate sensitivity.
From
is
s^^
the parameters of Eq. 4-9,
a fitted
it is
straightfon\'ard to calculate the parameters of the other possible representations.
To
<7
log.
where
in
4-9
calculate the strain rate sensitivity, the natural logarithm of Eq.
<t^^
—
1
The
ksi.
slope of the
fit is
= m log^, (£) - w log,,
(fy
is fit
to the data:
4-10
)
the strain rate sensitivity, w, and the intercept
is
the second term
Eq. 4-10.
Tables 4-1 and 4-2 summarize the
fitted
value of the strain rate sensitivity, m. In
tensile strength increase with increasing strain rate.
both F, and TS for each
steel.
Some of the
and round specimens for the quasi-static
used for each
test.
Appendix E provides
to evaluate the strain rate sensitivity.
and
Appendix
A
cases the yield and
all
contains strain rate sensitivity plots of
calculations utilize data developed using the standard Flat 2"
rates.
The
tables in
Figure 4-9
tensile strengths as a function of specified
Appendix
A identify which
combining the data from the
justification for
plots the
minimum
measured
strain rate sensitivity,
yield strength.
The
measured
in the longitudinal
Because the
total elongation,
is
no
specimens
m, for the yield
uncertainties in Fig.
Tables 4-1 and 4—2 are the estimated standard uncertainties of the calculated strain rate
described in the footnote of Table 4-1. There
specimen was
different
4-9 and
sensitivities,
w, as
significant difference in the strain rate sensitivity, w,
and transverse directions.
£/„ depends on the specimen geometry,
static test data to evaluate the strain rate sensitivity.
For
many of the
it is
not possible to use
core column
steels,
all
the quasi-
however, there
is
specimens of Fig. 4-1. Many of the quasi-static
0.25 in. Flat 2" specimen, which is geometrically
quasi-static test data obtained using the high-rate style
tests
on perimeter column
identical to the
justification for
?
= 0.125
steels
in.
employed the
/
=
Flat-HSR T'specimen. Appendix E provides
combining data generated using these two
theorefical
and experimental
specimens for evaluating the total elongation
behavior.
NISTNCSTAR
1-3D,
WTC
Investigation
91
Chapter 4
Calculation of Strain-Rate Sensitivity for Kolsky Tests
4.4.2
Figure 4-10 shows the strain rate dependence of the flow stress for the three steels evaluated in the
Kolsky bar
For the
A
tests.
For the
325 bolt
tv>'0
steel the
perimeter columns, the flow stress
flow stress
evaluated
is
at a 5
is
evaluated
percent offset
at a
1
percent offset strain.
Table 4-5 summarizes the
strain.
data plotted in Fig. 4-10. Table 4-6 compares the strain rate sensitivities evaluated in tension with those
evaluated in the compression Kolsky bar
The absolute values of the
tests.
strain rate sensitivities
evaluated by the two techniques are similar. The larger scatter of the Kolsky data, Fig. 4-10, results in
uncertainties in the calculated strain rate sensitivities, m, that are roughly
two times
larger than those
from
the tensile data.
0.07
—
1
'
1
I
1
1
1
'
1
1
1
'
'
'
'
1
Tensile strength:
-
o°c(de/clt)"'
0.06
o
A
TS
TS
longitudinal
transverse
[
^0.05
if)
0.04
B
0.03
c
CD
W
03
C
«
I
0.02
f
00
9
0.01
0.00
1
,
,
1
40
20
,
1
1
1
,
,
,
60
1
,
1
1
80
1
1
1
120
100
Fy, ksi
0.07
-i
1
r
Yield strength
0,06
O
A
w
c
0
w
0.04
B
0.03
c
'to
F,,
longitudinal
F,
transverse
9'
0.02
5^,
to
0.01
A
0.00
20
40
60
80
100
120
Fy, ksi
09 13:00
-
Thursday February
10,
2005
Figure 4-9. Strain rate sensitivity of yield and tensile
strength as a function of specified minimum
yield strength.
92
NISTNCSTAR
1-3D,
WTC
Investigation
High-Strain-Rate Properties
Figure 4-1
Note
in
that
1
by
dependence of the flow
plots the strain rate
5 percent strain the flow stress
is
stress at different strain levels for the
independent of the strain
Table 4-6, agrees with the strain rate independent strength found
rate.
A
325
bolt.
This behavior, summarized
in the full bolt tests
described in
Section 3.5.2.
Summary
Tabl e 4-5.
of stress-strain rate data plotted in Fig,
N8-C3M-FR2
Fy = 60
WTC
1
C22-C2M-FL3
Fy - 75 ksi
ksi
142 97-100
WTC
Strain
A
157 93-96
Strain
Unknown
Strain
Rate
Stress
Rate
Stress
Rate
(MPa)
(1/s)
(MPa)
(1/s)
(MPa)
(1/s)
500
0.001
650
0.001
990
0.0023
460
0.01
650
0.01
988
24
470
500
700
500
1020
384
550
750
800
600
968
1150
550
750
700
600
992
915
650
2400
700
1200
991
2970
575
1500
720
1200
650
2200
740
1250
620
2400
840
1750
620
2200
800
2400
550
3000
N8-C3M-FR2
325 bolt evaluated
Slow
and C22-C2M-FL3 evaluated
at 5
-10.
325 Bolt
Location
Stress
Notes:
A
1
^
at
1
percent offset strain.
percent offset strain.
rate tests evaluated using
method of Section
4.2.4.
Figure 4-10. Flow stress as a function of strain rate for Kolsky tests.
NISTNCSTAR
1-3D,
WTC
Investigation
93
Chapter 4
A 325
3.1
m
rr,
m
bolt
= -0.000010.0016 evaluated at e=0.05
= +0.0119+0.0041 evaluated at c:=0.03
= +0.0260+0.0061 evaluated at S:-=0.01
CO
CL 3.0
CO
O
2.9
CO
CD
13
^ 2.8
O
2.7
2
0
log^o (strain rate), 1/s
19:26.26
-
Monday December
20,
2004
Figure 4-11. Flow stress as a function of strain rate evaluated at different strains for
Kolsky tests on the A 325 bolt.
Table 4-6. Comparison of strain rate sensitivity measured
Bar and Compression
Tension Tests"
ksi
tension and compression.
Strain Rate Sensitivity from Kolsky
Strain Rate Sensitivity from
Specimen
in
C22-C2M-FL3
75
F,:
«;=0. 006710.0021
w=0.0114±0.0041
N8-C3M-FR2
60
Fy.
w=0.0 160100033
m=0.0152±0.0057
TS:
w=0.001 ±0.001
A 325
a.
TS
^ ^ mm
120 ksi
Bolt
Strengths evaluated at
1
Tests"
m=0.0000±0.0016
Stress evaluated at 5
% plastic strain
percent offset strain for both tension and compression.
HIGH-STRAIN-RATE DATA PROVIDED TO THE INVESTIGATION
4.5
Early in the Investigation, before the tensile testing was complete or
provide estimates of the high-strain rate behavior of the
These estimates were based on a provisional
set that
tlilly
analyzed,
it
was necessary
to
WTC steels to other Investigation team members.
contained only six perimeter-column and no core-
The provisional model predicted the increase in flow stress as a function of the strain rate
and the quasi-static yield strength, using the Cowper-Symonds equation. After the full set of high straincolumn
steels.
rate data
from the
thirteen steels
was complete,
the strain rate sensitivities for the
fiall
set
were analyzed
using the same methodology. The prediction of the high-strain-rate behavior of the perimeter column
steels
six
similar for the provisional and complete models.
perimeter-column
steels.
94
was generally
steels over-predicts the increase in
Appendix C summarizes
flow
The provisional model based on
stress in the low-strength
the
core-column
the analysis of both data sets.
NISTNCSTAR
1-3D,
WTC
Investigation
High-Strain-Rate Properties
COMPARISON WITH LITERATURE DATA
4.6
Table 4-7 summarizes the strain rate sensitivities of several structural steels for comparison with those of
the
WTC
steels in
Table 4-2.
It is
not a comprehensive
(Couque 1988. Krafft 1962, Manjoine 1944) and
structural steels.
The data from Chatfield
( 1
little
but
it
does include several plain-carbon
HSLA (Chatfield
974) and Davies
WTC
strucmral steels, but the chemistries are similar to the
(1962) and Couque (1988) provide
list,
( 1
1974, Davies 1975, Langseth 1991)
975 are for automotive rather than
)
steels.
Most of the
references, except Krafft
information on the microstructure of the steels characterized.
Table 4-7. Literature data for strain rate sensitivities of structural steels.
/;/
m
Determined
Determined
for Fy
Reference
Steel
1020
ir
for
TS
Chemistry
mass
%
(ksi)
(Eq. 4-9)
(Eq. 4-9)
Notes
28
0.043
n/r
(a) to
Not reported
30
0.059
n/r
(a) to
Not reported
31
0.054±0.007
0.019±0.005
(a) te
Not reported
40
0.045±0.004
0.031+0.007
te
Couque
1988
Couque
1020 T7'
1988
Commercial Manjoine
low-carbon
1944
open
hearth''
HSLA-40
Chatfield
1974
EM-3''
C
0.06
Mn
0.34 Si 0.02
Cu
0.04 Cr 0.03
Ni0.03 Al 0.055
Krafft
44.6
0.064
0.053
(b)c
45
0.035±0.003
0.03410.0009
te
C
0.20
Mn
1.12 Si 0.31
1962
HSLA-45-1
Chatfield
HSLA-45-2
Chatfield
45
0.024±0.002
0.023±0.005
te
30
0.026±0.003
0.01910.0001
te
50
0.014±0.002
0.01010.0018
te
51
0.021±0.002
0.009810.0005
te
1974
HbLA-50
LnatiieJd
1974
YST-50
Mn 0.78 Si 0.06 Cu 0.04
Mo 0.016 Al 0.001
C 0. 8 Mn 0.69 Si 0.63 Cu 0.07
Ni 0.12 Mo 0.012 Al 0.007
L 0.J3 Mn U.DD Si U.94 Lu 0.04
Ni 0.12 Mo 0.020 Al 0.012
C
0.18
Cr 0.03
Ni0.03
1974
Davies
1
CO.lOMn
Cr 0.03
Lr 0.42
0.54 Ti0.12
1975
ST52-3N
Langseth
Davies
Hot-rolled
C0.12
Ni 0.20
1991
60
n
r
Mn 1.30 Si 0.40 Cu 0.20
Mo 0.03 Al 0.03
0.02210.003
te
C0.19Mn
C
Cr 0;08
0.75
1975
HSLA-80-1
Chatfield
80
0.020±0.002
0.01610.003
te
80
0.018±0.007
0.01810.005
te
80
0.012±0.0007
0.008710.001
te
1974
HSLA-80-2
Chatfield
1974
YST-80
Davies
Mn 0.76 Si 0.60 Cu 0.03 Cr 0.49
Ni 0.03 Mo 0.005 Al 0.02
C0.05 Mn 1.55 Si 0.16Cu 0.01 Cr 0.02
NiO.Ol Mo 0.29 Al 0.068
CO.ll Mn 0.54 Ti 0.30 V 0.018
0.
14
1975
a.
Actual
b.
Reported value.
\
aiue forF,.
Key: te, tension test; to. torsion test; c, compression test; n/r = not reported.
Note: See Table 4-1 for definition of uncertainties of parameters.
NISTNCSTAR
1-3D.
WTC
Investigation
95
Chapter 4
Figure 4-12 plots the strain rate sensitivities, m, from Table 4-7, as a function of yield strength,
with the data from the
F,,,
along
WTC steels. The behavior and magnitude of the strain rate sensitivities of the WTC
steels is entirely consistent
with those from the literature of similar
steels.
0.07
Tensile strength:
ooc(de/dt)'^
0.06
WTC
LiteratLire
0.05
w
0.04
CD
CO
c
2
1 1
0.02
CO
I
t65
0.01
0.00
-I
20
40
I
J
L.
60
J
L.
I
I
1_
100
80
120
Fy, ksi
0.07
-|
1
1
1
r
1
1
Yield strength
0.06 -
G
vnc
•
Literature
0.05 -
w
0.04
CO
CO
1
'2 0.02
1?1
W
~9
0.01
0.00
—
I
40
20
60
100
80
120
Fy, ksi
09:05:55
-
Thursday February
10,
2005
Figure 4-12. Comparison of strain rate sensitivities of
NIST WTC steels to values for structural steels from
the literature.
One
limitation of the data
only
at the
AISI 1020
from the
extremes of the strain
steel at strain rates
WTC steels is that the strain rate sensitivity is determined using data
rates. Gilat et al.
near
1
s"'
was
aging. Other investigations of plain-carbon
(1997) found that the strain rate sensitivity of
negative.
and
They argued
HSLA steels
that
it
arose from dynamic strain
either did not probe this strain rate range
(Couque 1988) or did and found that the strength increases monotonically with strain rate (Chatfield
1974, Davies 1975, Krafft 1962).
96
Even
steels likely to
be susceptible
to dynainic strain
NISTNCSTAR
aging {Chatfield
1-3D,
WTC
Investigation
High-Strain-Rate Properties
1974. Krafft 1962. Manjoine 1944) exhibit this monotonic increase in strength with strain
to Gilat (1997).
Harding (1972) found three regimes of constant, but positive,
yield strength of BS 968, a ferrite-pearlite structural steel.
sensitivity
is
nearly zero. At rates up to
the strain rate sensitivity
There
is
is
this in\'estigation.
total
strain rates less than 0.1
is
w~0.07. At
the strain rate
s"'
rates
above
10"^ s"'
strain rate.
elongation and strain rate in structural steels in the strain-rate
Ahhough Bleck
(C = 0.05 percent) decreases with increasing
depend equally on
At
the strain rate sensitivity
of the
even higher.
no consistent relation between
range used in
10"^ s"'
rate. In contrast
strain rate sensitivity
(2000) found that total elongation, £/„ in mild steel
et al.
strain rate, they also
found
that values for other steels
do not
Manjoine (1944), Kendall (1965), Chatfield (1974), Davies (1975),
Langseth (1991). and Harding (1972) also found no strong correlation between
strain rate
and
HSLA
elongation. Bruce (2003) reported that ductility increases with increasing strain rate in
total
steels.
Figures 4-13 and 4-14 compare the strain rate sensitivities of the total elongation of various structural
steels
from the
literature to those
obtained in the Investigation. The data for the
WTC
steels
were
all
obtained using specimens of constant, identical geometry, but possibly different size, or using identical
Flat-HSR 1" specimens
at all rates.
The data
employed. The elongation behavior of the
tables in
steels
Appendix
A
identify the specific
from the perimeter and core columns
of the
specimen types
similar to the
is
behavior of other structural
steels. Uncertainties in the strain rate sensitivities
large for both literature and
WTC steels. In many cases, the measured strain rate sensitivities are
statistically indistinguishable
elongation are
from zero.
SUMMARY
4.7
The
total
strain rate sensitivity
of the yield and tensile strength of the perimeter and core column
similar to that of other structural steels reported.
The
total
generally increases slightly with increasing strain rate.
percent and
is
NISTNCSTAR
steels tested
is
elongation of perimeter and core column steels
The magnitude of the
increase
is
several absolute
similar to behavior reported for other structural steels.
1-3D,
WTC
Investigation
97
1
'
'
'
'
'
'
\
'
\
Total Elongation
50 ksi e,=(+1 .97±0.G0) M26-C1B-RF
F,,= 60 ksi e, =(+0.61 ±0.07) N8-C1B-RF
F^.= 100 ksi e -.;-f0,86±0.18) M10B-C3B-RF
F: = 100 ksi e.=(+1. 20+0.01) C10-C1M-LF
F^,=
50 -
UJ
EI,=e(j+e,logj(d£/dt)
40
30
20
All test
M26. N8 are
specimens have A
Flat
HSR
1",
/Lo
= 0.177
M10B, CIO are mixed
perimeter steels longitudinal orientation Fy>=50 ksi
10
I
I
I
I
I
-2
I
I
I
I
I
I
0
log^Q strain rate, 1/s
—————— —
1
A
50 _
'
'
'
I
'
'
'
F^=51
e,={-Q.00±0.71) ST52-3N
e,=!-0.60±0.84) HSLA50
e;=(-0,53±0 63) HSLA80A
e.,=(+1.08±0 30) HSLA80B
ksi
F.,=50 ksi
F^=80
F.=80
ksi
ksi
40
CO
§30
1-
20
Data sources
ST52-3N Langseth (1991)
HSLA... Chatfield (1974)
Literature Steels with
10
,
\
,
I
,
L
-2
I
I
I
I
f^>=50
I
I
ksi
I
0
log^o strain rate, 1/s
1
4:43:1 5
-
Thursday December 30, 2004
gure 4-13. Total elongation, Elt, as a function of strain
rate for high-strength, perimeter column steels and
high-strength steels in the literature.
NISTNCSTAR
1-3D,
WTC Investigation
High-Strain-Rate Properties
'
'
1
EI.=ep+e,logj(clE/clt)
50
F„=36
F,=36
r.=36
F„=42
0
•
lD
'
'
A
<-
'
\
'
1
Total Elongation
e,=(+0.83±0.27) B6152-2
e,=(+0.50±1 .37) C65-web
e.=!+0.62±0.56) C65-fiange
e.=(-0.19±0.02) HH-web
ksi
ksi
ksi
ksi
40
c
o
00
§
30
0)
o
20
All test
All test
10
I
HSR
1" with
= 0.125
/Lo
t=0.0625
in.
core steels longitudinal orientation F^,<50
ksi
I
-4
specimens have A
specimens are
I
I
Flat
I
I
-2
\
\
I
i_
I
0
log^o strain rate, 1/s
1
1
1
1
1
1
1
1
1
1
1
1
I
_
1
F„=29
F.=40
- 0
50 1
•
UJ 40
ksi
e.=(+l .31x0.39) Manjoine
ksi
e,={+0.90±1.09)
Ff.=45 ksi e.=(+0.52i:1 .89)
F;.=45 ksi e,=(+0.15±0.64)
HSLA40
HSLA45A
HSLA45B
—"—^— ——"u
-
-
-
^
c
o
O)
I
—
30
a
•
•
o
-
Data sources
20 h
—Manjoine
-
(1941)
HSLA... Chattield (1974)
Literature Steels with
10
1
1
-4
1
1
1
1
1
1
1
-2
1
1
F^<50
1
1
ksi
1
0
loQiQ Strain rate, 1/s
14 43.23
•
Thursday December 30, 2004
Figure 4-14. Total elongation,
E/,,
strain rate for low-strength, core
low-strength steels
in
as a function of
column
steels
and
the literature.
REFERENCES
4.8
Bleck,
W and
tests.
Bruce, D.
I.
Schael. 2000. Determination of crash-relevant material parameters
by dynamic
tensile
Archiv fur das Eisenhutlenwesen. 71 173-178.
M. 2003. Dynamic
Tensile Testing of Sheet Steels
Strengthening Mechanisms
NISTNCSTAR
1-3D,
WTC
in
and Influence of Strain Rate on
Sheet Steels. Ph.D. Thesis. Colorado School of Mines.
Investigation
99
Chapter 4
M. D. K. Matlock,
Bruce, D.
J.
G. Speer, and A. K. De. 2004. Assessment of the Strain-Rate Dependent
Tensile Properties of Automotive Sheet Steels.
SAE
SAE
Technical paper 2004-01-0507. Presented
at the
2004 World Congress. March 8-11.
Chatfield, D. and R. Rote. 1974. Strain Rate Effects on Properties of
High Strength,
Low
Alloy Steels.
Society of Automotive Engineers, pubhcation no. 740177.
Couque, H., R.
Asaro,
J.
Duffy, and
J.
S.
H. Lee. 1988. Correlations of Microstructure with
Quasi-Static Fracture in a Plain Carbon Steel. Met. Tram. A.
Dynamic and
19A .2179-2206.
Upon
the Tensile Deformation of Materials.
Dynamic Tensile
Tests," European Structural Integrity
Davies, R. and C. Magee. 1975. The Effect of Strain-Rate
JErig. Malei: Tech. 97[2]. 151-155.
ESIS Designation P7-00. 2000. "Procedure
KH. Schwalbe,
Society.
Editor.
for
Approved August 2000.
Follansbee, P. S. 1985. High Strain Rate Compression Testing,
Metals Handbook
9'''
ed.
Volume 8 Mechanical
Testing.
The Hopkinson
American Society
Bar. p.p 198-203. in
for Metals,
Metals Park,
Ohio.
Follansbee, P. S. 1985b. High Strain Rate Compression Testing, Introduction, in Metals
Volume 8 Mechanical
Gilat,
Testing.
American Society
Gillis. P.
P and
J.
1972. Effect of
Alloy Steels.
Harding
J.
steel
1987.
J.
High
Iron Steel
Testing.
over a wide range of strain rates and
Flow
American Society
Strain Rate
Inst.
ed.
13[6-7]. 611-632.
T. S. Gross. 1985. Effect of Strain Rate on
Volume 8 Mechanical
Edition,
Harding,
Int. J. Plasticity.
9'''
for Metals, Metals Park, Ohio. 190-192.
A. and X. Wu. 1997. Plastic Deformation of 1020
temperatures.
Handbook
Handbook Ninth
Metals Park, OH. p.p. 38-46.
Properties. Metals
for Metals.
on the Room-Temperature Strength and Ductility of Five
2 1 0. 425-432.
The Effect of High
Strain Rate
on Material Properties. Chapter 4 of Materials
at
High
Strain Rates. T.Z. Blazynski, ed. Elsevier. London.
Hutchinson,
J.
W. and K. W.
Neale. 1977. Influence of Strain-Rate Sensitivity on Necking Under
Uniaxial Tension. Acta Metallurgica. IS. 839-846.
M. and A. M. Sullivan. 1962. On Effects of Carbon and Manganese Content and of Grain
on Dynamic Strength Properties of Mild Steel. Trans. A.S.M. 55. 101-118.
Krafft,
J.
Langseth,
M. U.
ST-52-3N.
Manjoine M.
J.
S.
J.
Lindholm,
P.
Size
K. Larsen, and B. Lian. 1991. Strain Rate Sensitivity of mild steel grade
Eng. Mech. 117[4]. 719-731.
1944. Influence of Rate of Strain and Temperature on Yield Stresses of Mild Steel. Trans.
A.S.M. 66. A-2\\-A-2\S.
100
NISTNCSTAR
1-3D,
WTC
Investigation
High-Strain-Rate Properties
Wiesner, C.
S.
and H. MacGillivray. 1999. Loading Rate Effects on Tensile Properties and Fracture,
Proceedings of the Seventh Symposium Toughness of Steel, Fracture, Plastic Flow and Structural
Inlegiity: April 29, Abingdon, UK. 149-173.
Yan. B. and K. Xu. 2002. High-strain-rate behavior of advanced high strength steels for automotive
applications. 44" Mechanical Working and Steel Processing Conference Proceedings. Volume 40.
Association for Iron and Steel Technology. Warrendale, PA.
NiSTNCSTAR
1-3D, WTCInvestigation
101
Chapter 4
This page intentionally
102
left
blank.
NISTNCSTAR
1-3D,
WTC
Investigation
Chapter 5
Impact Properties
INTRODUCTION
5.1
Charpy impact
tests
were performed on
from the World Trade Center (WTC)
steels
comparison with the WTC-construction era
and
literature data,
to provide data for
to provide additional information
on how
mode of the steels. Estimates of the energy absorbed during an impact and
mode of the steel are both of interest. As strain rates increase, the strength of
strain-rate affects the fracture
the characteristic fracture
steel tends to increase,
impact testing
is
and the fracture behavior tends
one of the methods used
to
be
the impact tests with respect to the transition temperature
energy levels associated with a given
test
less ductile
to evaluate these effects.
from
and more notch-sensitive. Charpy
This chapter discusses the data from
ductile to brittle behavior, the absorbed
temperature, and the influence of specimen orientation on the
absorbed energy. The impact data generated in the Investigation were not used quantitatively used in the
modeling
efforts,
Figure 5-1
however.
a schematic
is
example of the definition of the
energy absorbed. At high temperature there
fracture
is
fracture
is brittle.
transition temperature expressed in terms
An
is
also relatively constant, but low,
such as 20
ft
and
S-shaped curve connects the two shelves. The transition temperature from ductile to
many
Common
different ways.
definitions based
on absorbed energy
include the midpoint of the S-curve, a specific fraction of the upper shelf energy, or
level,
of the
an upper shelf of relatively constant absorbed energy where
At low temperature the absorbed energy
ductile.
behavior can be defined in
brittle
is
Ibf (27 J) or
examining the change
in
30
ft
Ibf (41
The
J).
transition temperature can also
some
fixed energy
be characterized by
appearance of the fracture surface with temperature. The fracture appearance
transition temperature (or
FATT)
represents the temperature
at
which fibrous features occupy
fraction of the fracmre surface. In this chapter, transition temperatures based
a specific
on energy are defined as the
midpoint between the upper and lower shelf energies. Transition temperatures based on fracture
appearance are defined as the temperature
fracmre surface. For any
which ductile
at
fracture features
occupy 50 percent of the
temperatures measured by these two definitions will rarely
steel, the transition
be identical, but frequently they will be similar.
PROCEDURES
5.2
To
obtain the most useful data, representafive samples of the steels found near the impact zone were
evaluated. These samples do not represent the thicker secUons and the types of steels found in the lower
stories
of the towers. Impact
International
(ASTM)
test
specimens were oriented according
Standard Test Method
A
requirements of ASTM Standard Test Method
full-size (10
mm
x 10
mm)
(5
mm X
the
10
E
23-02.
specimen cross sections
too thin, and sub-size specimens were used.
to the
requirements of ASTM
370-97a. Tests were conducted according to the
The
to
Some components were
thick
enough
to
permit
be taken, but for many components the plates were
ASTM Type A specimen, with 5 = 5 mm thickness
mm cross section), was used for the sub-size specimens. Figure 5-2 shows the geometries of
two specimens.
NISTNCSTAR
1-3D,
WTC
Investigation
103
Chapter 5
The specimens were oriented according
to Section 4
of A 370. Figure 5-2
sample
illustrates the
orientations with respect to the rolling direction of the steel using a composite micrograph of the
microstructure of a plate from perimeter column
features are the bands of pearlite.
The
N9-C1TI-LF
(F,
=55
ksi). In the figure, the
is
such that the long dimension of the specimen
fracture direction
is
The longitudinal
is
parallel to the rolling direction,
and the
perpendicular to the rolling direction. The long dimension of the transverse specimen
is
across the plate or structural shape, perpendicular to the rolling direction, and the fracture direction
parallel to the rolling direction.
as
The
grains are the slightly elongated shapes with varied contrast.
inclusions are also elongated in the direction of rolling, but are not visible on this scale.
orientation
dark
L-T and T-L specimens
samples taken
at
is
These longitudinal (L) and transverse (T) specimens are often referred
to
respectively in other specimen orientation designation systems. For the
or near welds that join flanges and webs on perimeter columns, the welding direction
coincides with the rolling direction of the steels.
A transverse
specimen can be located with the notch
tip
adjacent to the fusion line of the weld and promote crack growth in a direction parallel to the weld
(parallel) in the heat affected
zone (HAZ).
All specimens were tested on a 300
E
23-02. The striker velocity
at
(Poussard 2003) to approach 5000
estimated to be about 100
s'',
ft
lbf capacity impact machine that
impact was approximately 17.7
s"'.
which
The average
is
ft/s.
was
strain rate across the
near that of the high-rate tensile
specimens, the standard anvil supports for 10
verified to the requirements of
Strain rates at the notch are estimated
ligament of the sample
tests.
When
is
testing sub-size
mm thick specimens were replaced with higher ones, so the
center of the striker met the center of the specimen.
Specimens were tested over
Because of the
scatter that is
a range
of temperatures
common
to characterize the transition
behavior of the
impact testing, multiple specimens were tested
in
temperatures. All impact energy data were collected digitally using a precision encoder.
surface appearance
was assessed
to estimate the fracture
which was defined as the temperature
features. Evaluations
at
which the
steels.
most
at
The
fracture
appearance transition temperature (FATT),
fracture surface exhibited 50 percent ductile fracture
were made using the chart comparison technique of ASTM E 23 Aimex A6. These
estimates of percent shear provide qualitative information, complementary to the absorbed energy data.
This investigation employed only one sub-size specimen geometry, so these results can be compared to
each other, but they cannot be directly compared to handbook data, which are usually developed on
standard, full-thickness specimens. There
size
specimen Charpy
tests
is
a large literature
using sub-size specimens. Lucon (2001) presents a concise
various methods for predicting the upper shelf energy,
temperature.
When
unlimited material
is
this
type of analysis
was not
of full-
summary of the
greater than the
is
correction factor
Charpy specimen
by analyzing both
for the Investigation
were
full-
less than 10
and sub-
mm thick,
possible.
For the scaling of energies, methods generally define a correction
energy obtained from sub-size specimen.
Ess, to
factor,
K, that scales the upper shelf
the expected energy from full-size specimens, Efs:
E,
104
results
as well as the ductile-to-brittle transition
available and plate thickness
many methods simply develop an empirical
specimens. Because many of the plates of interest
thickness,
size
on methods for predicting the
^KE
5-1
NISTNCSTAR
1-30,
WTC Investigation
Impact Properties
The
correction factors. K. take
include
many
different forms. Table 5-1
summarizes some
some permutation of the specimen width, 5, and ligament
depth,
b. Fig.
common
5-2.
More
ones. All
recent
correction factors incorporate details of the notch, including the notch radius, R, and notch angle,
The
6.
individual studies on specimen size effect (Corwin 1986,
Kayano 1991, Lucas 1986, Lucon 2001,
Louden 1988. Manahan 1990, Schubert 1995, Schwartzbart 1954, Uehira 2004) reach different,
incompatible conclusions about the suitability of the different correction factors. These discrepancies
probably arise because each investigator characterized different
steels,
using
test
specimens with different
notch and loading geometries. Several of the correction factors (Corwin 1986, Lucas 1986) do not
incorporate information about the notch geometry.
This investigation uses a correction factor, K.
that nearly all the
proposed by Lucas (1986)
^kMs ^ 10mmx(8mm)- ^
^
B^^b^
5 mm x (8 mm)"
^^
Note
first
proposed correction factors
in
Table 5-1 predict
K=2 because the full-size and
sub-size specimens have identical notch geometries. These scaling factors do not explicitly consider the
effect
of increased constraint for larger specimens, as
ASTM E 23-04
(2004) Appendix
XI
discusses.
Different scaling methods must be used for estimating the full-size specimen transition temperature from
sub-size specimens (Lucon, 2001).
The
transition temperatures in this report,
which were estimated from
fracture appearance, have not been scaled.
RESULTS
5.3
Appendix
A
of this report tabulates the data from individual impact
the important experimental results.
by location
The
rest
tests.
Table 5-2 summarizes some of
of this section summarizes the results for the
steels
arranged
in the building.
Perimeter Columns
5.3.1
The upper shelf energy of longitudinal specimens from flange material from perimeter column N8-C1M1
(Fy = 60 ksi specified strength, F, =68.8 ksi measured), located between the 97th and 100th floors of
WTC
1, is
mode
enters the transition region. For the transverse specimens, the absorbed energy drops
38.3
shelf of 10.4
ft
ft-lbf,
Fig. 5-3.
At about
-5 °F the absorbed energy begins to decrease as the fracture
Ibf beginning at about -40 °F.
that the transition temperature
(FATT)
is
about -95 °F. After correcting for sub-size specimen effects
using Eq. 1-2, the estimated upper shelf absorbed energy
longitudinal and transverse orientations respectively.
longitudinal to transverse orientations
The impact data
for the flange
of the
is
from an upper
For both orientations, fracture surface evaluations indicate
The
approximately
ClO-ClMl
is
76.6
ratio
ft
lbf and 20.8
ft
lbf for full-thickness
of the upper shelf energies for the
3.7.
perimeter column from
WTC
1
west face, column 452,
5-3 show that the upper shelf for both the longitudinal and transverse orientations
extends throughout the temperature range evaluated. Fracmre surface evaluations confirm this result and
floors 85-88 in Fig.
indicate that the transition
NISTNCSTAR
1-3D,
temperamre (FATT)
WTC Investigation
is
below -100 °F
for this
quenched and tempered
steel
105
Chapter 5
(specified
F,.
= 90
flange, there are
ksi,
supplied
two upper
F,.
= 100
shelves: 47.1
measured
ksi,
ft
F,.
=
109.4
ksi).
As
the case for the
is
lbf for longitudinal specimens and
1
1.4
ft
N8-C1M1
lbf for the transverse
specimens. After correcting for sub-size specimen effects, the upper shelf absorbed energies for full-size
longitudinal and transverse specimens are 94.2 ft-lbf and 22.8 ft-lbf respectively.
The
transition curve of steel
measured F, = 54.4
from the perimeter column spandrel N8-C1B1 (specified
5^,
ksi). Fig.
is
similar to that for the flange material from the
F,,
= 42
ksi,
same column. The
absorbed energy drops from the upper shelf at temperatures below about -40 °F. The upper shelves for
the longitudinal and transverse specimen orientations are 40.9 ftlbf and 13.3 ft-lbf respectively. Fracture
surface evaluations indicate the transition to 100 percent brittle fracture
is
nearly complete at -1 12 °F.
After correcting for the effects of the sub-size specimen using Eq. 1-2, the estimated upper shelf energy
for full-size longitudinal specimens
is
26.6
ft
is
81.8
ft lbf,
and the upper shelf energy
for the transverse specimens
lbf
HAZ
5.3.2
Materials from Perimeter
Columns
The upper-shelf absorbed energy values of HAZ (heat affected zone) specimens, notched near the toe of a
fillet weld on the flange of perimeter column N8-C1M1, are similar to those of the unaffected flange
materials. Fig. 5-3.
The upper
transverse specimens.
The
shelf
is
44.3
ft
lbf for the longitudinal specimens,
and 9.4
ft
lbf for the
data and fracture surface evaluations indicate that the specimens with
longitudinal orientations have entered the transition region at 0 °F, but the transverse specimens have not.
Again, after correcting to full-size using Eq. 1-2, the upper shelf absorbed energies for the longitudinal
and transverse specimens are 88.6
HAZ region
ft-lbf
and 18.8
ft
lbf respectively.
The
transition temperature
appears to be higher than the unaffected flange material, but
significance considering the transition temperature
HAZ
is still
quite
low and
this
has
likely
little
below 0
of the
practical
°F.
ClO-ClMl flange, tested using
transversely oriented specimens are similar to those of the unaffected ClO-ClMl transverse flange
material, Fig. 5-3. The average absorbed energy is 11. 0 ft-lbf down to temperatures below- 100 °F. After
The upper
shelf absorbed energies from the
material from the
correcting for sub-size specimen effects, the upper shelf absorbed energy for the transverse specimens
22.0
ft
is
lbf
Core Columns
5.3.3
Impact properties of specimens from the web of core column C-80, a
column 603
floors 92-95, (specified F,
= 36
ksi,
measured
F,.
=
14WF184 shape from
1
34.4 ksi) also vary as a function of
sample orientation. Fig. 5-4. The average absorbed energy of the room temperature
(53.8±6.3) ft-lbf for the longitudinal specimens, and (26.1±1.2)
WTC
ft
tests is
lbf for the transverse specimens.
The
absorbed energy decreases with temperature throughout the range tested, indicating transition behavior
over the range of 0 °F <
features at
1
106
172 °F. Fracture surface evaluations show approximately 95 percent ductile
room temperature. The transition temperatures (FATT)
room temperature and less than 140 °F.
ductile features at
than
r<
72 °F for both longitudinal and transverse specimen orientations, and less than 50 percent
for both orientations are greater
NISTNCSTAR
1-3D,
WTC Investigation
Impact Properties
Trusses
5.3.4
The data for the D = 0.92 in. diameter truss rod longitudinal samples (specified Fy = 50 ksi, measured
Fy = 60.0 ksi ) show a gradual transition in impact behavior throughout most of the temperature range
tested (Fig. 5-5).
The absorbed energy drops from 107
temperature the absorbed energy
ductile fracture at 172 °F
temperature (FATT)
The
steel
from the
Fy = 52.2 ksi-58.8
is
(32.4±4.8)
is
ft
lbf at 172 °F to 4.9
ft
lbf at -4 °F.
lbf Fracture surface evaluations
ft
and about 30 percent ductile fracture
at
room
At room
show about 90 percent
temperature.
The
transition
higher than room temperature, likely about 100 °F.
floor-truss angles (2
ksi) also
shows
in.
x 1.5 in.xO.25
in.)
(specified Fy
= 50
ksi,
measured
transition behavior throughout the temperature range tested (Fig. 5-5).
The absorbed energy drops from (31.1±1.1)
ft
lbf at
room temperature to 1.6 ft lbf at -76 °F. Fracture
(FATT) is likely below room temperature
surface evaluations indicate that the transition temperature
(50 percent to 70 percent ductile
room
at
using Eq. 1-2, the absorbed energy
3.2
ft
lbf
on the lower shelf at -76
5.3.5
Truss Seats
The Fy - 36
ksi truss seats
is
temperature). After correcting for sub-size specimen effects
(62.2±2.2)
ft
lbf at
room
temperature, (17.8±1.1)
ft
lbf at -4 °F, and
°F.
from perimeter columns,
like the floor-truss angle
and rod, show transition
behavior throughout the temperature range tested, Fig. 5-5. The absorbed energies for transverse
specimens range from 27.5
ft
lbf to 36.8
lbf at 172 F,
ft
temperamre the absorbed energies of truss
(1
1.9±0.5)
M4
ft lbf,
and (21.2+0.5)
ft
seats
at
about 2.3
77 °F, 50 percent ductile
The N8 samples were 20 percent
between room temperamre and 104
ft
lbf at
-40
°F.
At room
are (7.7±1.1)
°F,
at
ft
lbf,
N13 and
140 °F, and about 80 percent
ductile at 77 °F, 50 percent ductile at a temperature
and nearly 100 percent ductile
at
172 °F.
Bolts
5.3.6
At 68
to
lbf respectively. Fracture surface evaluations indicate that the
samples were about 10 percent ductile
ductile at 172 °F.
and drop
N13-C3B1, M4-C3T, and N8-C3T
°F. the
averaged absorbed energy for the
the fractures were 100 percent ductile.
The
A 325 bolt material
scatter
ft
lbf
As
the
temperamre drops
to
is
lower shelf energy
starts
somewhere below
5.4
DISCUSSION
5.4.1
Perimeter Columns
is
(41.1±12.4)
ft
lbf,
and
room temperamre absorbed energy
is
32 °F, fracture surface evaluations indicate the transition
entered. At -40 °F the fracture mode
temperature region
5-6
and scarcity of data prevent a more accurate estimate of
the average. Again, correcting for sub-size specimen effects, the
(82.2±24.8)
in Fig.
is
approximately 75 percent
brittle,
so the
this temperature.
Figures 5-7 and 5-8 summarize the impact toughness data for the various building components tested.
Data for sub-size specimens are normalized for comparison with data for
Eq. 1-2.
The summary
NISTNCSTAR
1-3D.
plots use the
WTC
Investigation
same
scales to facilitate
ftiU-size
specimens using
comparison.
107
Chapter 5
The
N8
transition temperatures of the
perimeter coluinn
steels. Fig.
5-7 and Table 5-2 are
in the
expected
range based on the relationships shown in Fig. 5-9 for the strength and toughness of HSLA steels in
1967 (Irvine 1967b). These data show that a fine-grain
ferrite-pearlite steel (or a controlled rolled steel)
with a strength 50 ksi <
to
< 65
F,.
ksi
would be expected
have a transition temperature in the range
-50°F<r<-100°F.
The impact data
quenched and tempered ClO-ClMl flange
for the higher strength
show no
Fig. 5-7,
indication of transition behavior
material, as well as the
HAZ
samples tested from
down
is
it,
-120
steel {Fy
=
109.6 ksi),
The transition temperature of this
well below -100 °F, which is far below the
to
°F.
temperatures predicted by the historic data in Fig. 5-9.
The
by
transition temperatures of the perimeter
historical data.
There
is
no reason
The upper
These
column
very tough
steels are
to expect that they
would
shelf energies of the perimeter
at
70
fail in
steels are similar
and are lower than those predicted
°F, the temperature at the time
a brittle
manner
at
high strain
of the
aircraft impact.
rate.
column longitudinal specimens, which have
a notch orientation
that is transverse to the rolling direction, are approximately three times larger than those of the transverse
by
the
By
mid 1970's this effect was understood (Irvine 1967, Wells 1975, Kozasu 1975,
Tamura 1988). The toughness of steel in the upper shelf region is influenced
morphology of inclusions, pearlite, grains and other constituents in the steel. Banded pearlite and
specimens.
DeArdo
the
1975, Farrar 1975,
increasing inclusion lengths, for example, were
known
to reduce the through-thickness (short transverse)
impact toughness of structural steels (Fig. 5-2).
One of the key
is
issues in understanding the effect of microstructure
that sulfide inclusions
become more
and approach the deformability of the
development of controlled rolled
had been
common
matrix
at
plate steels
with decreasing rolling temperatures
1650 °F (900 °C) (Baker 1972). In the 1960's, the
steel technologies resulted in the
in the past. This
was key, produced high
plastic relative to the steel
ferrite
on the toughness of these
use of lower rolling temperatures than
approach to plate production, for which a lower rolling temperature
strength, fine grain, ferrite-pearlite steels, with the adverse effect of elongating
the sulfide inclusions along the rolling direction. Since inclusions and other microstructural features that
can reduce toughness are oriented with their longest lengths and largest areas on planes parallel to the
plane of rolling, toughness can depend strongly on sample orientation in heavily rolled or controlled
rolled steels.
Figures 5-10 and 5-1
1
show
characteristically different,
When
that the fracture surfaces of longitudinal
which
the plane of crack growth
is
reflects the influence
and transverse specimens are
of the microstructure on the fracture process.
perpendicular to the rolling direction, the fracture surface
convoluted and has circular ductile dimple features where voids initiated
grew and coalesced just
prior to failure.
When
the direction of crack
direction, the fracture surface has ductile linear features,
peak-valley features, Fig. 5-12.
For
this orientation, the
at
manganese
growth
is
is
sulfide inclusions
parallel to the rolling
which form a corduroy texture with long
void growth
is
parallel
tubular rather than spherical,
because the lengths of the elongated inclusions are parallel to the crack growth direction. Fracture features
on the surfaces of these elongated ductile dimples.
Figs. 5-1 Ic
and 5-1
Id,
have morphologies
that
might
be expected for disrupted pearlite colonies and remnants of manganese sulfide inclusions. Figure 5-13
shows the locations and morphologies of manganese
sulfides
on the fracture surface of a transverse
impact specimen using an elemental x-ray map. These resuhs indicate that
108
many of the
NISTNCSTAR
film or plate-like
1-3D,
WTC
Investigation
Impact Properties
features
on the fracture surface are manganese
sulfides.
As
the
number and length of the manganese
sulfide inclusions in the steels increase, the resistance to crack growth in this orientation decreases.
The impact energies of these perimeter column flange and spandrel steels are in the range expected,
compared with data from the literature, see Table 5-3. The corrected upper shelf impact energies for
longitudinal specimens, for example, are about 75
ft-lbf,
which
is
the
similar to the 84 ft-lbf reported for a
0.012 S low carbon structural steel (Hulka 1993).
Although the upper shelf energies for the flange and spandrel
of the
steels
late 1960's. the ratio
steels are within the
of longitudinal to transverse energy
These differences
(Korchynsky 1970) and also
in the
control in steelmaking
significantly for
became common
practice
HSLA
uneconomical
cited as
HSLA
(Kozasu 1973).
in large scale steel
structural steels until the 1970's,
(Gladman 1997). As
content to a level below 0.01 mass percent, where
was
niobium X60
steels
impact toughness of longitudinal and transverse orientations were of concern
were not improved
the time, but
for heavily rolled
reported in
have been reported in WTC-era
literamre. Ratios of longitudinal to transverse properties of about 3
ferrite-pearlite steel
is
range expected for
near the maximum
late as
when
at
sulfur
1970 the reduction
in sulfur
transverse toughness begins to increase significantly,
production (Korchynsky 1970).
One of the common concerns for low through-thickness toughness was cracking under welds on structural
members, which was being discussed in relation to controlled rolled steels in the 1960's. The problem
was primarily viewed as a concern for thick sections, in which the high post-weld stresses could produce
under the welds due
large cracks
to delamination. In the case
of the
WTC perimeter columns and
spandrels, delaminations due to welding have not been observed to be a problem, but the lower transverse
toughness relates to the ductile tearing that occurred along the direction of rolling and the "pull-out"
failure
modes
at
welded components on the faces of the columns.
HAZ
5.4.2
Materials from Perimeter
The impact energies of the perimeter column
practical effect
orientation
is
HAZ
samples indicate that the welding operation had
on the toughness, Fig. 5-7. The lower impact toughness for the transverse
relevant to the fractures seen along the edges and
presumably due
5.4.3
Columns
at
the joints of
some
little
HAZ specimen
of the columns,
to overload during the collapse.
Core Columns
The absorbed energy at 77 °F of transverse web specimens from core column C-80, (26.1 ±1.2) ft lbf, are
similar to those that Barsom (1988) reported for slightly heavier wide-flange shapes. Table 5-3. In those
tests, the
12
<
average room-temperature absorbed energy was 37
CVN <
80
ft
lbf
The estimate
with the average value of
5.4.4
1
that the
13 °F that
FATT
is
Barsom (1988)
ft lbf,
but the data ranged over
between room temperature and 140 °F
is
consistent
reported.
Trusses
The impact energies
for the truss angle
and
rod. Fig. 5-8, increase with increasing temperature,
indicates transition behavior in the temperature range tested.
but fracture surface evaluations indicate that the
NISTNCSTAR
1-3D,
WTC
Investigation
FATT
is
The upper shelf is
near
which
not apparent in these data,
room temperature.
Literature data (Wilson
109
Chapter 5
1978) indicate that a range in transition temperature
A
expected for steel plate grades such as
of other
transition temperatures
A
36 and
(FATT) from about 70 °F
537, having thickness of 0.75
C-
late 1960's, 0.17
Mn
0.8
steels are also at or
(Duckworth 1967). Considering the nominal composition of the
C, 0.8
Mn, 0.004 N, 0.05
and 0.01
Si,
°F might be
The
to 1.5 in.
in.
above room temperature
WTC truss angle and rod (0.04 Al, 0.20
and the other factors known
to 0.03 S)
to 145
to influence the transition
temperature, such as pearlite content, precipitation strengthening, and free nitrogen content, only pearlite
content would be expected to raise the transition temperature higher than that predicted for the 0.17
0.8
Mn
factors
steels.
This
is
because the higher carbon content would increase the pearlite content,
remained unchanged. The nitrogen and silicon contents of the
truss steels are low,
C-
other
if all
and the sulfur
content affects only the absorbed energy and not the transition temperature. So, the near-room-
temperature transition temperatures of the angle and rod shapes from the trusses are consistent the
properties expected for this steel. Clearly, these materials
were
enough
ductile
which was
for fabrication,
the primary consideration.
Truss Seats
5.4.5
Figure 5-8 summarizes the impact data for the F,
N13, and M4. The
= 36
from perimeter columns N8,
ksi seat materials
above room temperature, and fracture
room temperature. In Fig. 5-14, for example, the
and M-4 samples show that fracture occurred in a predominantly brittle
transition temperatures for the truss seats are all
surface evaluations indicate very low ductility at
N- 1 3
fracture surfaces of the
manner
room
at
temperatures. This finding
may be
Expected Values of Impact Toughness
5.4.6
The measured impact properties of the recovered
context
their similarity to the properties
is
important to
to build the
require
relevant to possible scenarios for floor collapse.
remember
that there
WTC. No ASTM
Charpy impact
must be kept
of other structural
steels
were no specified requirements
a
minimum
current
AISC LRFD manual
of the
possess a
CVN
era.
For
this context,
on the
One
it is
steels
used
had been agreed upon
impact toughness of 15
in the late
ft
now,
(2001) only requires impact testing in one
CVN impact toughness of 20 ft lbf at 70
discussions of the era indicate that not even the simplest of
structural steels
proper context to be useful.
for impact properties
very limited case: steels for heavy Group 4 and 5 rolled shapes used
must have
in
specifications for structural steel intended for use in buildings, then or
The
tests.
steels
°F.
in tension
and spliced by welding
Furthermore, the literature
minimum toughness requirements for
One criterion, that the steel
1960's (Irvine 1967).
lbf at the operating temperature, originated with the
WWII
experience of the catastrophic failure of the Liberty ships. (Barsom 1987b)
Table 5-3 summarizes the historical data on Charpy impact toughness of structural
ASTM
specifications.
The impact
to the historical values.
A
The data
tougliness of
WTC steel evaluated at room
for core wide-flange
column C80 can be compared
36 wide-flange data in Table 5-3. The perimeter column plate steels
compared
to the
A
572 plate
properties similar to the
column
steels
A
exceeds 15
steels. Finally the F,
514 Type E
ft
steel.
lbf at 70 °F.
= 100
ksi perimeter
is
to the
quite similar
heavyweight
N8-C1M1 and N8-C1-B1 can
plate ClO-ClM has
be
column
The impact toughness of all
Only
steel supplied to
temperature
the impact toughnesses of
the of the historical
some of the
and
WTC
truss seats steels are
less than this value.
110
NISTNCSTAR
1-3D,
WTC Investigation
Impact Properties
SUMMARY
5.5
Table 5-2 summarizes the results of the Charpy
steels,
tests.
The upper
shelf energies of the perimeter
including material from the heat-affected zones of welds, average 76
19 ft-lbf transverse. Their transition temperatures are
temperamre
is
all
well below
room
ft
lbf longitudinal and
temperature, and
on the upper shelf The room temperature absorbed energies of the
wide-flange steels are about 40
ft
as 7
ft lbf,
e\'aluations
and
truss
their transition temperatures are all
room
and core-colunm
lbf longitudinal, and their transition temperatures are at
temperamre or above. The room-temperamre transverse absorbed energies of the
low
column
room
truss seat steels are as
above room temperature. Fracture-surface
demonstrated that the failure of the truss seat steels
at
room temperature was predominantly
brittle.
The impact
properties of the steels, except
contemporaneous structural
some of the
steels intended for
truss seat steels, are consistent with
use in buildings.
REFERENCES
5.6
AISC (American
of Steel Construction). 2001. Load and Resistance Factor Design Specification
Institute
for Structural Steel Buildings. Chicago,
II.
Section A.3.1c, page 3 and commentary on p.p. 168-169.
Sept. 4.
ASTM
(American Society for Testing and Materials). 2004. E 23-04 Standard Test Methods for Notched
Bar Impact Testing of Metallic Materials. Book of Standards volume 03.01. American Society for
Testing and Materials.
Baker, T.
J.
West Conshohocken,
Pa.
Use of Scanning Electron Microscopy
1975.
in
Studying Sulfide Morphology on Fracture
Surfaces. In Sulfide inclusions in steel: proceedings of an international symposium, 7-8
1974, Port Chester,
New
York.
J. J.
DeBarbadillo and
E. Snape, eds.
American Society
November
for Metals.
Metals Park, Ohio. 135-158.
Baker T.
J.
and
J.
A. Charles. 1972. Deformation of MnS Inclusions
in Steel.
Journal of the Iron and
Steel Institute. 210. 680-690.
M. and S. R. Novak. 1977. Subcritical Crack Growth and Fracture of Bridge Steels. National
Cooperative Highway Research Program Report 181. Transportation Research Board, National
Barsom
J.
Research Council.
Barsom
J.
M.
1987. Material Considerations in Structural Steel Design.
AISC Engineering Journal.
24[3]
127-139.
Barsom,
J.
M. and
S. T.
Rolfe. 1987b. Fracture
Fracture Mechanics.
Barsom
J.
M. and
2"^*
edition. Prentice-Hall, Inc.
New
in Structures.
Applications of
Jersey. Section 15.4.
B. G. Reisdorf. 1988. Characteristics of Heavy weight Wide-Flange Structural Shapes.
Welding Research Council Bulletin.
NISTNCSTAR
and Fatigue Control
1-3D.
WTC
Investigation
Number
332. 1-19. April.
111
Chapter 5
Corwin, W.R. and A.M. Hoagland 1986. Effect of Specimen Size and Material Condition on the Charpy
Impact Properties of 9Cr-lMo-V-Nb
Testing Irradiated Material,
for Testing
DeArdo
Jr.
A.
ASTM
Steel, p.p.
32-388in The Use of Small-Scale Specimens for
STP 888. W.R. Corwin and G.E. Lucas,
eds.
American Society
and Materials. Philadelphia, Pa.
J.
and E. G. Hamburg. 1975. Influence of Elongated Inclusions on the Mechanical
Properties of High Strength Steel Plate. In Sulfide inclusions in steel: proceedings of an international
symposium, 7-8 November 1974, Port Chester,
American Society
Duckworth, W.
New
York.
J. J.
DeBarbadillo and E. Snape, eds.
for Metals. Metals Park, Ohio. 309-337.
E. 1967. Metallurgy
of Structural
Steels: Present
and Future
Possibilities. 1967. In
Strong
tough structural steels: proceedings of the joint conference organized by the British Iron and Steel
Research Association and the Iron and Steel
1967. ISI Publication 104. Iron
Farrar
J.
M. and
C.
&
Institute at the
Steel Institute.
R. E. Dolby. 1975. Lamellar Tearing in
Inclusions. In Sulfide inclusions in steel: proceedings
1974, Port Chester,
New
York.
J. J.
Royal Hotel Scarborough, 4-6 April
London. 61-73.
Welded
Steel Fabrication.
The Role of Sulfide
of an international symposium, 7-8 November
DeBarbadillo and E. Snape, eds. American Society for Metals.
Metals Park, Ohio. 253-285.
Frank, K. H.,
J.
M. Barsom, and
R. O. Hamburger. 2000. State of the Art Report,
Fracture. SAC-2000-5a. Federal
Gladman,
T. 1997.
Low Carbon
Irvine, K.
J.
Emergency Management Agency.
The Physical Metallurgy of Microalloyed Steels. The
Hulka, K. and F. Heisterkamp.
Base Metals and
Low Carbon
Structural Steels,
Steels for the 1990's. R. Asfahani
The Key
Institute
to
Economic Constructions. 1993.
TMS.
and G. Tither, eds.
of Materials. London
211-218.
1967. The Development of High-Strength Structural Steels. In Strong tough structural steels:
proceedings of the joint conference organized by the British Iron and Steel Research Association and
the Iron and Steel Institute at the Royal Hotel Scarborough, 4-6 April 1967. ISI Publication 104. Iron
&
Steel Institute.
London. 1-10.
Kayano, H., H. Kurishita, A. Kimura, M.Narui, M. Yamazaki, and Y. Suzuki. 1991. Charpy Impact
Testing using Miniature Specimens and
Activation Ferritic Steels.
Korchynsky, M. and H.
J.
its
Application to the Study of Irradiation Behavior of Low-
Nucl. Mater. 179-181 425-428.
Stuart. 1970.
The
role of strong carbide
manufacture of formable high strength low alloy
Symposium jointly organized by
the Metallurg
steels.
and sulfide forming elements
Symposium low
Companies held
at
in the
alloy high alloy steels.
Nuremberg, BRD, on
May
21-23,
1970. 17-27
Kozasu,
I.
and
T.
Osuka. 1973. Processing Conditions and Properties of Control-Rolled Steel Plates. In
Processing and Properties of Low Carbon Steel,
New York.
112
J.
M. Gray,
ed.
The Metallurgical Society of AIME.
47-68.
NISI NC STAR
1-3D,
WTC Investigation
Impact Properties
Kozasu.
I.
and
J.
Tanaka. 1975. Effects of Sulfide Inclusions on Notch Toughness and Ductility of
symposium, 7-8
Structural Steels. In Sulfide inclusions in steel: proceedings of an international
November
1974, Port Chester,
New
York.
J. J.
DeBarbadillo and E. Snape, eds. American Society for
Metals. Metals Park, Ohio. 287-308.
Kumar.
Gamer, and M.L. Hamilton. 1993. Recent Improvements
A.S.. B.S. Louden, F.A.
Correlations for
DBTT and Upper Shelf Energy of Ferritic
Techniques Applied
to
Steels.
/?./>.
47-61
in
in Size Effects
Small- Specimen Test
Nuclear Reactor Vessel Thermal Annealing and Plant Life Extension.
STP 1204. W.R. Corwin
et al., eds.
American
ASTM
Society for Testing and Materials. Philadelphia, Pa.
Louden, B.S.. A.S. Kumar, F.A. Gamer, M.L. Hamilton, and W.L. Hu. 1988. The Influence of Specimen
Size on Charpy Impact Testing of Unirradiated HT-9.
J.
Nucl. Mater. 155-157. 662-667.
Lucas, G.E., G.R. Odette, J.W. Sheckherd, P. McConnell, and
V-Notch Specimens
,
ASTM
STP
888.
Perrm. 1986. Subsized Bend and Charpy
305-324in The Use of Small-Scale Specimens for
for Irradiated Testing.
Testing Irradiated Material
J.
W.R. Corwin and G.E. Lucas,
eds.
American Society
and Materials. Philadelphia, Pa.
for Testing
Lucon, E. 2001. Material Damage Evaluation of Primary Power Plant Components Using Sub-size
Specimens. Advanced Engineering Materials. 3[5]. 291-302.
Manahan. M.P. and C. Charles. 1990.
A Generalized Methodology for Obtaining Quantitative Charpy
Data from Test Specimens on Nonstandard Dimensions. Nucl. Tech. 90. 245-259.
C, C. Sainte Catherine, P. Forget, and B. Marini, 2003. On the Identification of Critical
Damage Mechanisms Parameters to Predict the Behavior of Charpy Specimens on the Upper Shelf
Poussard,
Predictive Material Modeling:
Combining Fundamental Physics Understanding, Computational
Methods and Empirically Observed Behavior.
STP
M.
T. Kirk
1429. American Society for Testing and Materials,
and M. Erickson Natishan,
West Conshohocken,
Impact Properties of Neutron Irradiated
H
and
J. P.
I.,
Uehira A. and
1
Size on the
Steels.
ASTM Proc.
54 939-963.
C. Ouchi, T. Tanaka, and H. Sekine. 1988. Thermomechanical Processing of High Strength
Low Alloy
of
ASTM
Sheehan. 1954. Effect of Specimen Size on Notched Bar Impact Properties of
Quenched-and-Tempered
Tamura,
eds.
Pa.
M.L Hamilton. 1995. Effect of Specimen
A533B Steel. J. Nucl. Mater. 225. 231-237.
Schubert. L.E. A.S. Kumar, S.T. Rosinski, and
Schwartzbart,
in
Steels.
S.
Butterworth
&
Co. Ltd. London.
Ukai. 2004. Empirical Correlation of Specimen Size Effects in Charpy Impact Properties
lCr-0.5Mo-2W, V. Nb Ferritic-Martensitic
Stainless Steel.
J.
Nucl. Sci. Tech. 41 [10] 973-980.
Wells, R. G. 1975. Metallographic Techniques in the Identification of Sulfide Inclusions in Steel. In
Sulfide inclusions in steel: proceedings of an international symposium, 7-8
Chester,
New York.
J. J.
November
1974, Port
DeBarbadillo and E. Snape, eds. American Society for Metals. Metals Park,
Ohio. 123-134.
NISTNCSTAR
1-3D,
WTC Investigation
113
Chapter 5
Wilson A. D. 1978. Comparisons of Dynamic Tear and Charpy V-notch Impact Properties of Plate
Trans.
114
ASME.
Steels.
100. 204-211.
NISTNCSTAR
1-3D,
WTC Investigation
Impact Properties
Common
Table 5-1.
sub-size to full-size upper shelf energy correction factors.
K
Correction Factor,
Reference
K
Note a
Definitions
B = specimen
Convin 1986
B
2
ss
bss
width
b = ligament depth
(see Figure 5-2)
Corwin 1986
2.8
Lucas 1986
bX
Louden 1988
Bjsbl
IbJI
bX IbjI
lX I lX
Schubert 1995
2
L = span = 40
2
2
K,
= notch
K,
'
=
mm per Fig. A 1.2 of ASTM E 23
stress concentration factor''
alternate notch stress concentration factor
a.
K is evaluated
b.
K, and K, depend only on the notch angle, 9. root radius, R, and ligament depth,
for the
NIST Charpy specimen
geometries.
'
full-size
b.
and sub-size specimens of the Investigation. Formulas for AT, can be found
Note: Subscripts 55 and
NISTNCSTAR
fs refer to
1-3D,
WTC
These three parameters are
in
Kumar
identical in the
(1993).
sub-size and full-size specimens, respectively.
Investigation
115
Chapter 5
Summary
Table 5-2.
of
Charpy
data.
Eus
Transition
longi-
trans-
T
E77
tudinal
verse
(FATT)
longitudinal*^
E77
ftlbf
Specimen
Description
ksi
flange
60
N8-C1M1
/=0.3125
Location"
WTC
1
(Note
°F
ftlbf
transverse''
(Note b)
(Note d)
ftlbf
ftlbf
c)
142 97-100
76.6
20.8
-95
76.7±2.9
21.6+0.3
452 85-88
94.2
22.8
<-100
92.0±2.4
23.3±0.7
142 99
81.8
26.6
n/d
83.6±2.2
26.4±0.8
142 97-100
88.6
18.8
<0
87.6±6.2
19.2±0.4
<-100
n/d
23.1±1.2
53.8+6.3
26.1+1.2
in. plate
sub-size specimen
ClO-ClMl
WTC
100
flange
/=0.25
1
plate
in.
sub-size specimen
N8-C1B1
WTC
42
spandrel
1
7=0.375 m. plate
sub-size specimen
N8-C1M1
Heat-affected
n/a
WTC
1
zone
sub-size specimen
ClO-ClMl
Heat-affected
n/a
WTC
1
452 85-88
n/d
22.0
36
WTC
1
603 92-95
n/d
n/d
11<T<
zone
sub-size specimen
WF web
14WF184
C80
core
full-size
truss rod
truss angle
specimen
Z)=0.92 m.
full-size
2
140
50
unknown
n/d
n/d
^100
32.4±4.8
n/d
50
unknown
n/d
n/d
<11
62.2+2.2
n/d
140
n/d
7.7+1.1
T < 104
n/d
21.2±0.5
140
n/d
11.910.5
< T<32
n/d
n/d
specimen
in. X 1.5 in.
xO.25 m.
sub-size specimen
N13
truss seat
full-size
N8
truss seat
full-size
M4
A
325 bolt
WTC
1
130 99-102
n/d
n/d
36
WTC
1
142 97-100
n/d
n/d
77 <
specimen
truss seat
full-size
36
specimen
36
unknown
n/d
n/d
n/a
unknown
82.2
n/d
specimen
D=7/S m.
bolt
-40
sub-size specimen
a.
Location code example;
WTC
142 97-100: Tower
1
1.
column
line 142,
between floors 97 and 100.
temperature evaluated from fracture surface appearance (FATT).
upper shelf absorbed energy, corrected using Eq. 1-2 where appropriate.
b. Transition
c.
d.
£77
=
=room temperature absorbed
Key: HAZ, heat affected zone;
n/a
=
energy, average and standard uncertainty, corrected using Eq. 1-2 where appropriate.
long, longitudinal orientation; trans, transverse orientation.
not applicable; n/d, not determined.
Note: The reported uncertainty,
is
the estimated uncertainty of the
mean of n
values:
1
n{n -
116
1)
NISTNCSTAR
1-3D,
WTC Investigation
Impact Properties
Table 5-3. Historical data on Charpy impact toughness of structural steel
CVN
T
Year
Steel
Type
ftlbf
°F
Orient
Barsom and
Novak 1977
1977
A3 6
plate
28
72
L&T
Barsom and
1988
Reference
A? 6
wide-flange
37
70
A3 6
1994
Range: 12
A572 Gr 50
ft
lbf<CVN<80
ft
lbf
Ib/n
wide-ilange
95
91%
70
shapes
1972
3
28 individual shapes.
W14 shapes
>342
Barsom 1987
single plate t=] in. Figure C-1 and
Table
Reisdorf 1988
Frank 2000
Notes
plate
of the specimens had
CVN>30ftlb
55
70
L
@70
°F, Table 6-4
also appears in (Frank
2000)
52 plates from 6 mills
Barsom 1987
1972
A572 Gr 50
plate
37
40
L
Barsom 1987
1972
A572 Gr50
plate
21
0
L
Barsom 1987
1972
A572 Gr50
plate
25
70
T
Barsom 1987
1972
A572 Gr 50
plate
18
40
T
Barsom 1987
1972
A572 Gr 50
plate
12
0
T
Barsom and
1988
A572
wide-flange
38
70
T
W14
Reisdorf 1988
>342
Frank 2000
A572
1994
22 individual shapes
Range:
shapes
11
wide-flange
61
77%
70
of specimens had CVN>30ftlb
r,
Barsom and
Novak 1977
1977
Barsom and
Novak 1977
1977
A514 Type E
ft-lbf<CVN<81 ftlbf
lb/ft
65
plate
70
L
1
dole D-H
single plate,
?
=
1
in.
Figure C-10 and
Table 3
A514 Type E
plate
40
a.
Year
b.
CVN:
c.
Orient: specimen orientation with respect to rolling direction
70
T
refers to first year of publication for references that cite other sources.
absorbed energ>'
in
Charpy
test.
Key: L, longitudinal: T, transverse.
NISTNCSTAR
1-3D.
WTC
Investigation
117
Chapter 5
80
— — — — — — — — — — — — — — — — — — — —— —
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
\
I
I
I
\
\
|'
\
'
'
'
'
I
transition region
Upper
Lower
I
I
shelf
shelf
I
-150
I
I
J
I
-100
-50
I
I
1
I
I
I
I
1
I
100
50
0
L
150
200
T, °F
11 02:01
-
Tuesday February
15.
2005
Figure 5-1.
An example
transition curve.
Specimen Orientation
rolling direction
ASTM A 370 longitudinal
impact specimers
L-T orientation
ASTfvl
A 370 transverse impact specimen
T-L orientation
Specimen Geometry
55
mm
B=5
b=8
55
mm
mm
,
I
mm
,
,
.
!
i
II
B=10
b=8
mm
„
0
mm
ASTM E 23 Type A
sub-ssze
speamen
mm
T
numm
ASTM E 23 Type A
fyji-size
^ecimen
Figure 5-2. Charpy impact specimen geometries and orientations with respect to the
plate rolling direction.
118
NISTNCSTAR
1-3D,
WTC
Investigation
—
—
—
r
Impact Properties
60
T
——— —— — — — — —— —— —
^
I
I
I
I
I
I
I
I
\
I
I
'
'
I
I
\
'
'
I
I
C10-C1M1
F =100
50
o
§
o
Q
©
a
i
1=—
»= — 1
O
40
§
ksi
WTC1 452
0
O
C10-C1M1
t
C10-C1M1 T
85-1
L
h
30
o
CD
O
20
<
.
10
•
"=
ClO-C1f/n
HAZ T
subsize specimen
J
>
I
I
I
I
I
I
I
-100
-150
I
I
i_
-50
50
0
100
150
200
T, °F
60
— — —TT— — — — —
T
I
I
I
I
I
—n— — — —
I
I
I
I
I
I
I
I
r
N8-C1M1
F =60
ksi
WTC1
142 97-100
50
NS-CIWil-HAZ L
40
>.
O)
Q
N8-C1M1-FL
H
o
NS-CIMI-FL T
L
oi
o
30
o
CD
o
w
20
<
•1^1=
10
N8-C1M1-HAZT
L=longitudinal; T=transverse
subsize specimen
0
I
-150
-100
I
-50
I
I
I
I
i_J
50
0
I
L.
100
150
200
T,°F
10:05:35
•
Tuesday March
15,
2005
Figure 5-3. Longitudinal and transverse Charpy impact data of samples from the flange
and adjacent HAZ of perimeter column N8-C1M1, the flange and adjacent HAZ of
perimeter column C10-C1M1.
NISTNCSTAR
1-3D,
WTC
Investigation
119
Chapter 5
60
——————— —— — —— — —— — —
T
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
\
N8-C1B1 -spandrel
F =42
ksi
WTC1
142
'
50
40
N8-C1B1-spandrel L
h
>.
/
CD
/o
S 30
D
CD
O
20
CO
<
/.
10
I
I
J
1-
1
i
N8-C1 El -spandrel
subsize specimen
_l
I
I
-150
I
I
I
I
I
-100
L
I
....
-50
J
I
I
\
I
I
100
50
0
L_J
I
I
150
L
200
T, °F
150
T— — — — — — — —r-|— — — — —[1
I
I
I
I
I
I
I
I
I
I
c80-web
F =36
ksi
WTC1
603 92-95
100
cn
CD
c
CD
D
CD
O
50
CO
<
T=transverse
:80-web L
L=longitudinal
full-size
I
-150
,
-100
,
,
specimen
Q8p-y.-Qb,T,it-
-50
50
0
100
150
200
T, °F
15:13:17
-
Friday March 11.
2005
Figure 5-4. Longitudinal and transverse Charpy impact data for the spandrel associated
with perimeter column N8 and the web of wide-flange core column C-80.
120
NIST NCSTAR
1-3D,
WTC Investigation
r
Impact Properties
150
——— ———————— ——
T
I
I
1
I
I
I
I
I
I
I
I
I
Truss seats
F =36
ksi
M4
N13
N8
100
0)
c
<D
D
CD
O
50
CO
<
I
-150
-100
=?
-50
full-size
r-,-1-
r.-^r
0
I
I
50
100
I
,
I
specimen
I
I
I
150
I
I
200
T, °F
Figure 5-5. Transverse Charpy impact data from samples from perimeter column truss
seats M4, N13, and N8, and from floor truss components.
NISTNCSTAR
1-3D,
WTC
Investigation
121
Chapter 5
60
— — — — —— — — —
-1
I
I
I
I
I
I
I
I
l'
'
'
'
'
I
A 325
50
O
bolt
O
40
>.
D)
0)
o
30
-o
0)
X3
O 20
CO
<
10
subsize specimen
-150
-100
_L
-50
I
0
I
I
50
100
,
I
I
I
I
I
150
I
I
200
T, °F
18 09,15
-
Tuesday Oclober
05,
2004
Figure 5-6. Longitudinal Charpy impact data for
122
A
325 bolts.
NISTNCSTAR
1-3D.
WTC
Investigation
Impact Properties
Figure 5-7. Summary plot of the dependence of absorbed energy on test temperature for
all perimeter and core column steels. The absorbed energy values of the sub-size
specimens have been corrected using Eq. 5-2 to compare them to data from full-size
by 10 mm) specimens.
(10
mm
NISTNCSTAR
1-3D.
WTC
Investigation
123
Chapter 5
T—
— — —TT—
I
120
I
—
I
1
I
I
I
I
——
————— — — — —— — —
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Truss seats
F =36
100 -
ksi
«
M4
A
N13
N8
m
80
>.
D)
i_
<D
C
60
0^
X3
40
§
<
§/
I-"
20
^^^
_i
I
I
I
I
I
I
—
—
-50
—
|—
Truss components
I
I
I
I
F =50
I
I
I
~T"
I
I
I
-100
-150
120 -
i
I
I
I
E_i
I
I
full-size
I
I
50
0
I
I
I
I
specimen
I
100
I
I
L
I
I
200
150
—— —— — — ——— ——— —— — —— — ———
I
T
-^A
I
.i'"
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
ksi
D=0.92
rod
in,
100
>;
80
c
0^
o
o
60
/
'^,2
iry,
X
1
.5 in.
angle
O
i
40
<
20
0
0
I
-150
I
-100
PrTT
-50
corrected
I
50
0
,
I
I
to full-size
I
I
I
100
specimen
I
I
I
150
I
I
200
T, °F
16:28:08
-
Tuesday October
05.
2004
Figure 5-8. Summary plot of the dependence of absorbed energy on test temperature for
all truss component and truss seat steels. The absorbed energy values of the sub-size
specimens have been corrected using Eq. 5-2 to compare them to data from full-size
by 10 mm) specimens.
(10
mm
124
NISTNCSTAR
1-3D.
WTC
Investigation
Impact Properties
50
1
1
1
1
1
1
1
1
1
1
/
1
1
\
/
0
_
/
Fine Grained
1
1
—
/.'^
Bainite,^^
Quenched and
Tempered
/ Fentte- /
_
-
^"zy^
Controlled
Rolled
/Controlled^^
/Rolled
c
1
1
/
/
/
I
1
Bainite
/
/
Ferrite-Peaiiite
-50
-100
A
1
1
/ Pearlite/
-150
-200
J
50
1
1
1
1
60
1
1
1
1
1
70
1
1
1
1
1
1
1
1
80
1
1
90
1
1
1
1
100
Yield Strength, ksi
Figure 5-9. Strength-toughness relationships for several types of structural steels from
the WTC construction era, after Irvine (1969).
c
Source: NIST
Figure 5-1 0. Scanning electron micrographs of the fracture surface of a Charpy V-notch
longitudinal specimen orientation from an N8-C1M1 perimeter column (WTC 1-142-97100). (a) ductile dimples (oval features) and general surface morphology, (b) low
magnification view of large and small ductile dimples on the fracture surface, (c) higher
magnification view.
NISTNCSTAR
1-3D,
WTC Investigation
125
Chapter 5
Source: NIST.
Figure 5-11. Scanning electron micrographs of the fracture surface of an N8-C1M1
perimeter column sample showing ductile tearing features that form due to fracture
initiation and growth at elongated inclusions and pearlite on planes parallel to the rolling
plane. The "ductile dimples" in this case are linear features with a peak-valley
morphology
Figure 5-12. Perspective view of the fracture surface of sample N8-C1M1 showing the
long peak-valley features characteristic of the fracture surface for transversely oriented
impact specimens. The green line indicates the topography of the fracture surface.
126
NIST NCSTAR
1-3D,
WTC
Investigation
Impact Properties
C
D
Source: NIST.
Figure 5-13. A gray-scale image (a) and compositional maps from fracture surface of an
N8-C1M1 perimeter column Charpy V-notch specimen. The relative concentrations of (b)
iron, (c) manganese, and (d) sulfur. The surface of the "ductile dimple" is littered with the
remnants of manganese sulfide inclusions.
Source: NIST.
Figure 5-14. The fracture surface of a perimeter truss seat, N13-C3B1 that was tested at
room temperature shows cleavage facets, which indicate a brittle fracture mode.
NISTNCSTAR
1-3D.
WTC
Investigation
127
Chapter 5
This page intentionally
128
left
blank.
NIST NCSTAR
1-3D,
WTC
Investigation
Chapter 6
Elevated Temperature Properties
INTRODUCTION
6.1
The high temperature
finite
tensile
deformation and creep behavior of structural steels are important inputs to
(WTC)
element models of the response of the World Trade Center
segment of the Investigation used two approaches
structures to the fires. This
to provide that information.
The
was
first
to
experimentally characterize the high-temperature mechanical properties of the steels that were most likely
to
have been exposed
to the fires.
The second was
to
develop methodologies to estimate the tensile and
creep properties of those steels that were not experimentally characterized.
The experimental
were most
characterization of the high-temperature tensile properties focused on the steels that
likely to
have been exposed
the core columns, and the inner
web
of creep properties focused on the
The methodology
to estimate
to high temperature
from the
The experimental
plates of the perimeter columns.
steel
the floor trusses and truss seats,
fires:
characterization
from the floor trusses
high-temperature tensile properties of untested steels recognizes that the
yield and tensile strength of structural steels, normalized to their
room temperature
values, degrade with
temperature along a master curve. To estimate the dependence of the shape of the stress-strain curves with
temperature, the methodology uses both literature and experimental data.
For estimating creep properties of all
methodology uses 1970s
strengths.
WTC structural steels, except those in the floor trusses, the
literature data for creep
For the properties of the
of structural
truss steels, the
steel scaled
by
the ratios of their tensile
methodology uses data generated from
tests
on
recovered truss steels themselves.
6.2
TEST procedures
6.2.1
Tensile Tests
High-temperature tensile
tests
employed two
employed an electromechanical
12.5
testing
mm gauge length. The furnace
Flat C2,
conform
long and 3
to Fig.
1
is
different machines.
a split, MoSii-element design.
of ASTM International
mm x 6 mm wide.
The
machine (Instron 8562) with
(ASTM) E
They were loaded using
21.
tests
on the
floor truss steels
a contact extensometer that
The specimens, designated type
The uniform cross section
pin-aligned, superalloy,
wedge
specimen temperature was monitored using a K-type thermocouple mounted within
surface.
The specimens were loaded
temperature, which
had a
1
is
grips.
32
mm
The
mm of the specimen
as soon as possible after the furnace temperature stabilized at the test
was generally within 20 min. During
the temperature stabilization step, the specimen
temperature was always within 5 °C of the desired temperature. The crosshead speed in these
tests
was
0.0325 mm/s, which produced a strain rate after yielding of approximately 0.001 s\ measured using the
extensometer. Prior to yield, the crosshead displacement rate produced a specimen stressing rate of
4±2 MPa/s. These
NISTNCSTAR
rates
1-3D,
meet the requirements of ASTM E
WTC
Investigation
8
and E 21.
129
Chapter 6
Most other
1
.0 in.
tests
used a second electromechanical machine with
gauge length. The furnace
specimens. Type Flat CI, have a
was usually
is
1.1 in.
a
high-temperature extensometer with a
which quartz lamps heat the specimen. The
a split design in
uniform long cross section 0.25
the original thickness of the plate.
These
tests
wide. The specimen thickness
in.
followed the loading protocol of ASTM
crosshead displacement rate was 0.00167 mm/s. After several percent
The
initial
rate
was increased
stress for elevated
to
0.0167
required by
irmi/s, as
temperature
E
strain, the
21.
crosshead
21, which produces a discernible step in the flow
tests.
Creep Tests
6.2.2
Creep testing employed two types of machines. Some employed the same electromechanical
with MoSi2 furnace used in the high-temperature tensile
Most
tests.
traditional dead-load, lever-ann creep frames (Applied Test
NiCr
E
employed two
tests
Systems) with
split,
test
identical,
wire-wound, three-zone
furnaces. Extensometers in these furnaces are averaging, clamp-on types with a nominal
mm. The
length of 25.4
specimens were loaded by hand by adding masses
took less than three minutes to reach
full load.
During the
test,
gauge
weight pan. Generally,
to the
the operator
machine
removed mass
it
to maintain
constant stress from the weight pan at approximately 3 percent strain increments. These corrections
manifest themselves as kinks in the strain-time curves. Specimen C132-2w-2 shows such a correction
approximately 6,000
were conducted
at
s
(as
shown
later in Fig. 6-3).
The
test
specimens were the type Flat C2. Creep
temperatures 400 °C, 500 °C, 600 °C, and 650 °C.
was used per specimen. Individual specimen
were generally discontinued
after 2
h
if the
One
stresses lay in the range 100
specimen had not
failed,
at
tests
stress-temperature condition
MPa < a < 445.8 MPa.
though some
tests
Tests
were allowed
to
run longer.
6.3
RESULTS
6.3.1
Tensile Tests
Table 6-1 describes the specimens from the three perimeter column, two core column and two floor truss
steels tested at
which
illustrate data
curves for
In
high temperature and their original locations in the buildings. The curves of Fig. 6-1,
all
from specimen N8-C1B1 A-FL,
the specimens tested.
A contains the
stress-strain curves for the other
most cases, the experimenter removed the extensometer prior
do not necessarily represent
E
Appendix
are typical of the high-temperature stress-strain
21, there
is
a
jump
in the
failure.
Because some of these
flow stress
at
tests
6.3.2
it
can be as large as 5
followed the loading protocol of ASTM
about £ = 0.02 that corresponds to the mandatory extension rate
change. At low temperatures, where the strain rate sensitivity
temperatures,
is
low, the
jump
is
small, but at elevated
ksi.
Creep Tests
Truss steel from a 2
in.
for creep. This steel
was
x 1.5
in.
x 0.25
in.
truss top chord angle
specified to conform to
A
242 with F,
from specimen C-132 was characterized
50
ksi.
Chemical and mechanical
analyses of the larger lower chord angles, which were specified to conform to
were made from the same
130
specimens.
end points of the curves
to failure, so the
steel,
A
36, indicate that they
however.
NISTNCSTAR
1-3D,
WTC
Investigation
Elevated Temperature Properties
Table 6-1. Specimens and locations for high-temperature tensile tests with
full
stress-strain data.
Specimen
Description
Source
(ksi)
C40-C2M-IW-2
Perimeter coluntn inner
N8-C1B1A-FL
Perimeter column flange
ClO-ClM-FL-1
Perimeter column flange
C65
12WF161
HH-FL-1
12WF92
core column
WTC
WTC
WTC
WTC
WTC
C-132
2
inx0.25in. truss bulb
Unknown
50
Unknown
36
in. X 1.5
web
core column
1
Column 136
floors 98-101
60
1
Column 142
floors
97-100
60
1
Column 451
floors 85-88
1
Column 904
floors
1
Column 605
floors 98-101
86-89
100
36
42
angle
C-53
3 in.
x2 inxO.37
in.
truss bulb
angle
Note:
F, is the specified \'alue. rather
than the NIST-measured value.
Figure 6-1. Elevated-temperature stress-strain curves.
Specimen N8-C1B1A-FL is from a Fy = 60 ksi perimeter
column flange plate from WTC 1 column 142 between
floors 97-100. Annotations refer to individual test
specimen numbers.
NISTNCSTAR
1-3D.
WTC
Investigation
131
Chapter 6
The creep
test
matrix consisted of 20 creep tests taken to failure or discontinued after several hours.
Figures 6-2, 6-3, 6-4, and 6-5 display the creep curves for specimens from the truss upper chord.
t,h
0.0
0.5
1.0
~1~
—
—
—
—
0.05
—
n
'
~i
'
'
650 °C-
7w-1 147 MPa
6n-2 145.4 MPa
2n-3 126 MPa
2w-8 125 MPa
c
ffl
0.04
/
D)
CD
CD
C
0.02
discontinued-
LU
tailed
creep test
0.01
—
fit
[/
'
0.00
J
I
I
I
I
I
I
I
I
I
I
Thursday October
-
I
I
I
I
I
I
I
I
L
I
4000
3000
5000
s
t,
19;23 07
I
\
2000
1000
0
2004
14,
Figure 6-2. Creep curves of A 242 truss steel from specimen C-132 at 650 °C. Dashed
lines represent the fit from Eq. 6-14 using the parameters in Eqs. 6-16, 6-17, and 6-18.
Experimental curves are graphically truncated at £ = 0.05.
t,
0.0
0.5
1.0
2.0
1.5
3.0
2.5
—m— — — — — — — — — — — — — — — — — — — — — — — — — — — — — —
I
"T
h
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
n
I
I
I
I
600 "C
discontinued"
tailed
creep
test
-
fit
5w-2 164 MPa
4w-6 150 MPa
3w-3 151.8 MPa
2VV-2 131.4
1
w-2 123
'
[
MP£
MPa
.
1.0
xlO'
19:231 6
-
Thursday October
14,
2004
Figure 6-3. Creep curves of A 242 truss steel from specimen C-132 at 600 °C. Dashed
fit from Eq. 6-14 using the parameters in Eqs. 6-16, 6-17, and 6-18.
Experimental curves are graphically truncated at s = 0.05.
lines represent the
132
NISTNCSTAR
1-3D.
WTC Investigation
—
Elevated Temperature Properties
t,
h
1
1
1
r
2.0
1
500
1^
'C
discontinue3"
failed
creep
test
-
fit
1n-2 300 MPa
2n-2 296.8 MPa
4W-.S. 277.9 MPa
lw-4 272.2 MPa
252 MPa
250 MPa
3w-2 200 MPa
3n-l
4v»,'-3
0
2000
4000
t,
-
6000
S
1923 19 Thursday Oclober 14 2004
-
A 242 truss steel from specimen C-132 at 500 °C. Dashed
from Eq. 6-14 using the parameters in Eqs. 6-16, 6-17, and 6-18.
Experimental curves are graphically truncated at e = 0.05.
Figure 6-4. Creep curves of
lines represent the
fit
Figure 6-5. Creep curves of A 242 truss steel from specimen C-132 at 400 °C. Solid lines
represent measured creep strain. Dashed lines represent the fit from Eq. 6-14 using the
parameters in Eqs. 6-16, 6-17, and 6-18. Experimental curves are graphically truncated
at e = 0.05.
NISTNCSTAR
1-3D.
WTC
Investigation
133
Chapter 6
6.4
MECHANICAL PROPERTY VALUES FOR STEELS
6.4.1
A
Universal Curve for Elevated-Temperature Tensile Properties
Early in the Investigation, before any significant tensile testing was completed,
provide other
of structural
NIST
Investigation
steel as a fiinction
Team members
all
of the different grades, the relation
WTC
were not characterized. The
particularly relevant for
Because
steels.
is
literature contains
The small
modeling the
it
was not experimentally
These were primarily the
fire-affected floors.
WTC construction era. A
and often recognized relation
at
elevated temperature.
however, because
strong disadvantage of the sources
of structural
steel as a function
yield strength drops
F,.,
and
of parameters
tensile, TIS, strength
away from
plateau out to about 300 °C.
unity
it
that
of temperature,
normalized to the room-temperature value, follows a universal curve, independent of strength
the normalized yield,
is
in the tests.
that the strength
is
steels in the core
were also characterized.
steels
very few summaries of the properties of structural steels
they do not typically report the strain rate used
feasible to
temperature behavior were those that were
data set that is available is particularly appropriate to the Investigation,
comprises structural steels from the
A useful
to
necessary to estimate the properties of the steels that
steels tested for elevated
columns and the floor tmsses, though several perimeter column
The
was necessary
of temperature. This relation forms the basis for estimating the elevated-
temperature stress-strain behavior of the recovered
characterize
it
with a relation to describe the yield and tensile strength
level.
Both
curves are roughly sigmoidal. The normalized curve for
more quickly than does
the curve for tensile strength,
which has
a
A function that represents this behavior with a smooth curve and a minimum
is
/=
FAT)
=
(1
- A2)e\p
6-1
F,(23°C)
JJ
and
TSjT)
(1
- ^Oexp
6-2
TS{23°C)
where 7 is measured
in "C.
The parameters
in Eqs.
6-1 and 6-2 are completely empirical and have no
physical significance.
Figures 6-6 and 6-7 plot the master curves for yield and tensile strength overlaid on the literature data,
which
USS
are the small solid
symbols (Chijiiwa 1993, Goda 1964, Harmathy 1973, Holt 1964, Melloy 1963,
The model curves, Eq. 6-1 and 6-2, are not fits to the literature data. Instead, the parameters
of Eqs. 6-1 and 6-2 were adjusted to provide a visually accurate representation of the bulk of the data,
particularly in the region of steep slope in the range 500 °C < r< 700 °C. Furthermore, the values of the
1972).
parameters, summarized in Table 6-2, have no physical significance. The scatter in the literature data
arises
from several sources. At intermediate temperatures, the
dynamic
strain
aging
differ.
Dynamic
strain
sensitivities
of the different
steels to
aging can actually raise the flow stress above the room
temperature value. Furthermore, the literature sources do not report the testing rates employed, which
134
NISTNCSTAR
1-3D,
WTC
Investigation
Elevated Temperature Properties
may
differ
between the sources. Rate
stress at temperatures
effects
can make a significant contribution to the measured flow
above about 500 °C.
Figure 6-6. Ratio, f, of room- to high-temperature yield strength (Fy) for all steels
characterized. The data spread at room temperature exists because the individual test
results are normalized to their mean. The solid line is the expression, Eq. 6-1, developed
using literature data on structural steels, which are denoted by the smaller symbols.
1.0
o
O
CM
0.8
-
o
0.6
C/)
CO
0.4
Model
.
» literature da'.a
core truss seats
0 core columns
•
0.2
A
truss
components
perimeter truss seats
o perimeter columns
0.0
200
400
T,
600
800
°C
Figure 6-7. Ratio of room- to high-temperature tensile strength (TS) for the steels in
Table 6-1. The data spread at room temperature exists because the individual tests are
normalized to their mean. The solid line is the expression developed for literature data on
structural steels, Eq. 6-2, denoted by the smaller symbols.
NISTNCSTAR
1-3D,
WTC
Investigation
135
Chapter 6
Table 6-2. Values for the parameters
in the strength reduction equations
(Eqs. 6-1 and 6-2).
Yield Strength
Reduction Parameters
Tensile Strength
Reduction Parameters
(Eq. 6-1)
(Eq. 6-2)
A. = 0.074692
A2 = 0.086327
m, = 8.325929
«,
m. =
H2= 4.992176
531.403687
/, =
.000000
1
T
5,= 638.384277
S2 = 523.523132 "C
The
recent Eurocodes
(EN 1993-1-2
/.=
Fire)
recommended
stress-strain
recommend
T
T
1032.864014
expressions for the proportional limit,^ and flow
of temperature. These two parameters define specific points on
stress, /max,of stRictural steel as functions
the
= 5.0004
behavior for
steel at elevated
temperature in the Eurocode. That curve has
four regions:
1
An
a small region of linear-elastic behavior
.
up
2.
a work-hardening region that extends to e
3.
a plastic regime that extends at constant,
4.
a failure range
to the proportional limit, /,,
=
0.02,
maximum
where the strength decreases
important feature of this stress-strain curve
is
that
it
flow
stress,
/^nax,
and
to zero.
reaches the
maximum
stress, /nax. at e
=
0.02.
Figure 6-6 compares both of these to the Investigation data. Neither the Eurocode proportional limit,^
nor the e = 0.02 yield strength,
/i^nax,
corresponds to the e = 0.002 offset yield strength reported in
Figure 6-6, but they do represent lower and upper bounds. There
is
no analogous definition
for the tensile
strength, TS.
Analysis of Tensile Data
6.4.2
Figures 6-6 and 6-7 also summarize the normalized yield (F,) and tensile (TS) strength for
by component. At temperatures up
tested, segregated
in the
same range
as the literature data.
to
The
strain rates
used for the
WTC steels generally lie
WTC data average only 75 percent of
WTC data were as low as
data were probably generated using rates appropriate to
ASTM A
£ = 8.3 X
two
lO'^'s"'
.
The expected difference between
that the high-temperature flow stress, o,
depends on the
cr
where £„
is
a constant and
The necessary
model curve
136
is
m
is
=
£"
=
from the differences
8.3 x
in the
0' s" while the literature
370, which are ten times higher:
testing rates can
strain rate
1
£"
in a
be evaluated by recognizing
power-law
relation:
6-3
£,£'"
designated as the strain rate sensitivity.
strain rate sensitivity,
w = 0.13.
the
the steels
400 °C the data from the
At 500 °C and above, the
the curve that represents the literature data. This deviation probably arises
testing rates.
all
w, to raise the average of the
Reported elevated-temperature
WTC data to the prediction of the
strain rate sensitivities for
NISTNCSTAR
low-carbon structural
1-3D,
WTC Investigation
Elevated Temperature Properties
steel are
near this value:
m = 0.105
at
600 °C (Manjoine 1944),
analysis of the strain rate sensitivities of six
extension rate
is
changed
at
WTC
steels,
m=
from the
0.08 at 649 °C
stress
jump
about 2 percent strain yielded similar values. The strain rate sensitivity,
evaluated from the individual strain rate jumps, increases with temperature:
777
= 0.119±0.020
600 °C, and
at
77?
= 0.146±0.029
at
m
Estimating Elevated-Temperature Stress-Strain Curves
A complete
stress-strain curve as a function
it
was not
= 0.055±0.029
at
/»,
500 °C,
650 °C.
6.4.3
modeling. Because
that
(Gowda 1978). An
occurs when the
of temperature
is
necessary as input for finite element
feasible to characterize all twenty different
WTC
steels for their high-
temperature stress-strain beha\ior. a methodology was necessary to estimate the stress-strain curves for
each
steel.
The
stress-strain (a-s) curves
two
of the more than twenty different
minimum
WTC structural steels can be
The shapes of the a=
8 curves of low yield strength (A 36, F,
36 ksi) steels differ from those of higher yield strength
{Fy > 36 ksi) steels. The low strength steels work harden over a larger strain range, even when they
grouped
into
classes based
on
shape and specified
their
develop the same tensile strength {TS) as the higher strength
At elevated temperatures, of course, the yield and
steels initially increase in strength
generated
dynamic
steels.
tensile strengths decrease, see Figs.
6-6 and 6-7. Some
with increasing temperature, through the process of dynamic strain-
aging, (Baird 1970) but this behavior
the effects of
yield strength.
strain aging.
not a priori predictable. The analysis does not attempt to predict
is
The limited quantity of literature
in the Investigation, indicates that the
stress-strain curves, as well as those
shape of the work-hardening portion of the stress-strain
curve changes with increasing temperature. The model to describe the stress-strain curves must capture
this
shape change.
The methodology
1
.
for creating stress-strain curves for
Choose
a representative low-strength
steel for
which
literature or
WTC steels has three steps.
(A 36
F,.
= 36
ksi)
and a high-strength
experimental stress-strain curves
at different
(F,.
> 36
ksi)
elevated
temperatures are available.
2.
De\ elop
a
model
that can predict the stress-strain behavior
of these
steels as a function
of
temperature.
3.
Apply
this
model
stress-strain
behavior
to all other
WTC
steels
by appropriately scaling the
resulting stress-strain behavior of the steel in question to that of the steel used to develop the
model.
The
stress-strain curves for 0
°C <
r<
650 °C
for all steels are represented using a
power-law work-
hardcnmg model:
a=
Rj^R^K{T)£"^'^^
6-4
where a has units of ksi, Thas units of °C and
NISTNCSTAR
1-3D,
WTC
Investigation
137
Chapter 6
KiT) = {k,-k,)exp
+
Vk2
6-5
k,
J
and
/
"I
=
ii{T)
^
A"
6-6
(«4 -/?o)exP
V '»2 y
The parameter
room
i?rs is
the ratio of the
room temperature
of the
tensile strength
steel
of interest, TSi to the
,
temperature tensile strength, TSreu of the steel used to develop the model expressed by Eqs. 6-5
and 6-6.
TS,
6-7
Because the work hardening decreases with increasing temperature, scaling by the room-temperature
tensile strength, TS,
is
more appropriate than
scaling
by
the room-temperature yield strength,
The
F,,-
parameter Rc corrects for two additional phenomena. The temperature-dependent functions (Eqs. 6-5
and 6-6) were originally developed using data
that
stress-strain curves supplied to the Investigation
zero strain
rate.
at
to zero strain rate,
Secondly, because the temperature dependence of all the possible steels
using the behavior of only two different
Eq. 6-4
were not corrected
room temperature could
steels,
differ
from
temperature stress-strain curve. The parameter
it is
model curve.
In general, the correction
is
represented
by the already generated and supplied room-
corrects the elevated-temperature stress-strain curve so
r= 25 °C at 8 = 0.05
small: 0.9 < Rc< 1.04.
of stress predicted by Eq. 6-4 for
is
to the
possible that the stress-strain behavior predicted by
that predicted
Rc
that the value
whereas the
teams for room temperature behavior are corrected
equals that of the
room temperature
The functions K(T) and n(T) have the same form as the functions that describe the reduction in yield and
6-6 and 6-7. Although they are not based on a physical model,
tensile strength with temperature in Figs.
they represent the behavior of n(T) and K(T), and are relatively easy to evaluate. Their functional form
produces a monotonic decrease
the trend of the literature data.
To
in the strength
of the work hardening with temperature, which agrees with
The maxima of the
functional forms of n(T) and K(T) occur at
calculate the six parameters in each expression fox
test stress-strain curve, i.e. the strain at stress greater
K(T) and n(T), the
plastic portion
7=
0 °C.
of each tensile
than the yield strength, was modeled with a power-
law work hardening expression.
cj
= Ke"
6-8
using a non-linear least-squares fitting method, to estimate an individual Ki and
n,
were then
fitted
maximum values
with Eqs. 6-5 and 6-6 respectively. The
for n /
and
ri:
,
which were necessary
fits
for n(T)
n,
.
The individual
were subject
to avoid introducing
to constraints
anomalous behavior
A',
and
on the
in
Eq. 6-4 at intermediate temperature. Note also that the functional form of Eqs. 6-5 and 6-6 does not
capture the small strength increase caused by dynamic strain aging
138
at
temperatures less than 300 °C. In
NISTNCSTAR
1-3D,
WTC
Investigation
Elevated Temperature Properties
addition to depending sensitively on testing rate, the magnitude of this increase varies from steel to steel.
It
may
be absent
many
in
steels, particularly the
one of the modeling groups required
with temperature, which
The model
higher strength, microalloyed
stress-strain behavior in
this functional
for the low-strength steel stress-strain curves
u,
to
which Eqs. 6-5 and 6-6 were
parameters in Eqs. 6-5 and 6—6 for steels with
fit.
= 36
F,.
is
steel.
based on the
A
36 data of Harmathy
model
K, and
n, in
Figs.
( 1
970).
Table 6—4 and Figs. 6-8 and 6-9 show the
Table 6-5 summarizes the values of those
ksi.
Figure 6-10 shows the prediction of the model
plotted o\ er the original data used to generate the parameters for Eqs. 6-5 and 6-6.
some of the
Computationally,
form supplies.
Table 6-3 summarizes some of the properties of this
individual Ki and
steels.
which the strength decreased monotonically
6-8 and 6-9 deviate strongly from the
Note
that although
fitted curves, the prediction
of the
for the stress-strain behavior. Fig. 6-10, agrees with the data quite well.
Table 6-3. Property data for the
A
36 steel reported
in
Harmathy
(1970).
A 36
Grade
C
Composition
0.19%
%
S
0.03 %
Mn 0.71 %
Si
0.09 %
P
0.007
semi-killed, from 1/2
in.
compositions e.xpressed
plate
in
mass percent
(20 °C)
44
ksi (reported, not corrected to zero strain rate)
TS (20 °C)
64
ksi (reported, not corrected to zero strain rate)
Range of data
20 °C - 649 °C
Number of tests
10
Table 6-4. Individual K, and
n,
used
for steels with Fy = 36 ksi.
T
NISTNCSTAR
1-3D.
WTC
°C
ksi
24
92.95
0.1404
99
82.2
0.1335
149
108.8
0.1947
204
140.0
0.2458
260
155.9
0.2766
315
152.6
0.2928
427
105.3
0.2352
«,
535
54.98
0.1402
593
38.41
0.1316
649
24.81
0.1167
Investigation
139
Chapter 6
— — — — — — — — — — — — — —r~i—
150
)
I
I
I
I
I
)
I
I
I
I
I
I
I
'~i
T
—
I
f
100 -
50
A
K measured
individual test
in
K(T)
J
I
I
I
I
I
I
I
I
I
I
I
I
200
100
I
I
L.
I
Monday October
-
11,
I
I
I
I
I
i
I
I
500
I
I
l_
1
I
700
600
°C
T,
15.58.08
J
400
300
2004
Figure 6-8. K(T), Eq. 6-5, for the A 36 steel of Harmathy (1970), used to model the
behavior of steel with F„ = 36 ksi.
— — — — — — — — — — — — — —nn— — — — — — — — — — — — — — — — — —
0.30
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
.
I
I
I
I
I
I
I
,
I
I
,
I
I
0.25
0.20
c
A
0.15
A
A
0.10 -
0.05
A
0.00
,
n measured
n(T)
I
0
I
,
,
I
I
I
in
I
200
100
I
individual test
I
,
I
I
I
I
300
T,
1
5:58:1
1
-
Monday October
1 1
Figure 6-9. n(T), 6-6, for the
140
.
I
400
500
600
700
°C
2004
A
36 steel of Harmathy (1970) used to model the behavior of
steel with Fy = 36 ksi.
NISTNCSTAR
1-3D.
WTC Investigation
Elevated Temperature Properties
Table 6-5. Values of the parameters of
Eqs. 6-5 and 6-6 for steels with
= 36 ksi
.
29.0492 ksi
hi
524.1812 °C
tk2
523.6799 °C
k,
9.4346
k:
9.3532
k.
121.6056 ksi
nil
0.1235
tnl
524.4304 °C
521.2410 °C
a
19.0000
19.0000
0.2168
n4
a.
0
Fits
on
/?/
and
were constrained
/7j
I
n<
19.
I
0.10
True
-
have
I
0.05
0.00
16 59 05
to
I
,
I
I
0.15
I
0.20
strain
Monday February 07 2005
Figure 6-10. Predictions for the model (dashed lines) for A 36 steel (Fy = 36 ksi nominal)
overlaid on the original data used to generate the model (solid lines). The model
(Eqs. 6-4, 6-5, and 6-6 and Table 6-4) makes essentially identical predictions for
0°C < T < 300 °C, so only one line is plotted. Note that the model should not be used for
strains in the elastic region (e~ < 0.003), but the curves are shown in this region. Instead,
elastic lines of the appropriate modulus should be used.
The model
for the high-strength steel stress-strain curves
laboratory of steel
steel.
The Laclede
steels. Its
truss steel,
summarized
in
1-3D,
in
Table 6-6
WTC
is
a
1
WTC
Investigation
and
in the
NIST
WTC 2 (location in
good model
composition and mechanical behavior are similar to an
Chemical and mechanical characterization indicates
NISTNCSTAR
based on measurements made
taken from the truss angles C-132 and C-53, used
building unknown).
high strength
is
for
most of the
ASTM A 572 Grade 50
that although the truss angles
of C-132 were
141
Chapter 6
supplied to
A
to A 36, the steels are identical, highASTM A 572. Most of the perimeter steels in the
242 and the tmss angles of C-53 were supplied
strength, low-alloy
(HSLA)
steels similar to current
towers are also micro-alloyed
steels, especially
those up to about F,
Table 6-6. Property data for the
A 242
= 60
ksi.
Laclede truss steel tested as part
of the Investigation.
A
Grade
Composition
c
0.19%
Mn
0.82
P
0.010%
S
0.029
Ni
Cr
0.10%
Cu
0.32
V
0.038
from 2.5
ksi)
%
%
0.07 %
0.08 %
Si
242 (F, - 50
%
%
in.
x 1.5
in.
x 0.25
in.
open web floor
truss angle
compositions expressed in mass percent
F, (20
"O
59 ksi
(static F, -corrected to
zero strain rate)
r5(20"C)
77.8 ksi (static ZlS-corrected to zero strain rate)
Ranee of data
20 "C - 650 "C
Number of tests
14
Table 6-7 and Figs. 6-1
and 6-12 show the individual Ki and
to which Eqs. 6-5 and 6-6 were fit.
Table 6-8 suirmiarizes the values of those parameters in Eqs. 6-5 and 6-6 for steels with F, > 36 ksi.
1
Figure 6-13 shows the prediction of the model plotted over the original data used to generate the
parameters for Eqs. 6-5 and 6-6. The model captures the shapes of the stress-strain curves. The actual
data
show
the reduced ductility characteristic of the Laclede truss steel.
Table 6-7. Individual
T
rc)
and
used
n,
for steels with Fy > 36 ksi.
K
(ksi)
n
25
117.52
0.1276
25
122.9
0.1699
25
113.6
0.1141
300
137.4
0.1763
400
400
142
K,
95.48
111.3
0.1460
0.1614
500
64.72
0.0815
500
68.12
0.0833
600
35.32
0.0415
600
30.16
0.0267
600
3
0.0250
600
37.03
0.045
650
23.0
0.0327
650
27.46
0.0353
1
.0
NISTNCSTAR
1-3D,
WTC
Investigation
Elevated Temperature Properties
150
100
50
0
200
100
0
400
300
T,
16:39 11
Figure 6-1 1
.
-
Monday Oclober
11,
500
600
700
°C
2004
K(T), Eq. 6-5,for the
A
242 Laclede steel used to model the behavior of steel
with Fy > 36 ksi.
0.30
0.25
0.20
c
0.15
0.10
0.05
0.00
0
200
100
300
T,
16 39 18
-
Monday Ociobei
11
400
500
600
700
°C
2004
Figure 6-12. n(T), Eq. 6-6,for the
A
242 Laclede
steel,
used to model the behavior of
steel with Fy > 36 ksi.
NISTNCSTAR
1-3D.
WTC
Investigation
143
Chapter 6
Table 6-8. Values of the
parameters of Eqs. 6-5 and 6-6 for
steels with Fy > 36 ksi.
IbAlli
ko
c
ksi
Ml.OiOD or^
L
hi
hj
1
1
O'") ii/i
1
1
OHIO
o/^
1
ki
O.DV
1
Zz. J
i
/
O
/
KSI
0.0342
'hi
519.6340
499.603
a
"C
1
10.0000
"/
10.0000
0.1511
a.
100
fits
on
n<
10.
T
1
;?/
and
z?!
1—
1
were constrained
to
I
1
1
1
1
1
have
1—
^
1
1
1
1
25°C
-
80
•
w
60
(/)
CD
to
^
40
20
0
1—1
0.00
1
I
I
0.05
I
I
I
I
I
True
16 55 48
-
Monday Oclober
1 1
,
I
I
0.10
I——I
I
I
0.15
^
I
I
0.20
strain
2004
Figure 6-13. Predictions for the model (dashed lines) for steel with Fy > 36 ksi overlaid
original data used to generate the model (solid lines). Note that the model should
not be used for strains in the elastic region (e- < 0.003), but the curves are shown in this
region. Instead, elastic lines of the appropriate modulus should be used.
on the
144
NISTNCSTAR
1-3D,
WTC Investigation
Elevated Temperature Properties
A small disadvantage of this method is that the very low strain (8 < 0.001) behavior is neither Unear nor
does
it
strain
ha\'e a slope equal to the
beha\
ior. this is a ver>'
for small strains
is
Young's modulus appropriate
small error, however.
to solve Eq.
to that temperature, hi
One way of introducing
6-4 for the intersection
which can be evaluated using Eq. 1-2 of the Chapter
point,
e,,,
terms of the stress-
linear stress-strain behavior
with the correct Young's modulus, E(T),
1
R^sRc^(T^)^r"'^' = EiT)£,
For
£
<
£y
use the linear Young's modulus relation; for s > Syuse Eq. 6-4. Figure 6-14 shows stress-strain
cur\ es calculated using the
method of Eq. 6-9. Figure 6-15 compares
method with the expression
Fig. 6-6.
6-9
The agreement
is
method. Further accuracy
for the yield point as a function
the yield points calculated
of temperature for
many
by
that
structural steels.
within the band of the scatter of the data, confirming the suitability of this
not warranted, because as temperature increases the yield strength depends
is
increasingly strongly on the testing rate. In these cases, the proportional limit
may be much
less than the
0.02 percent offset yield strength.
0.000
0.010
0.005
0,015
0.020
True Strain
1
7 34;S1
'
Monday Octobei
1 1
,
2004
Figure 6-14. Simulated elevated temperature stress-strain curves for the Laclede A 242
truss steel. The small-strain behavior is modeled using the appropriate Young's
modulus, while the large-strain behavior comes from Eq, 6-4.
Generally, the correspondence between the original
predicted by Eq. 6-4
is
quite good.
For the
steels
room temperature stress-strain curves and those
F,, > 75 ksi, the agreement is less satisfactory
with
because these steels do not strongly work harden. During the development of this methodology
could not locate any elevated temperature stress-strain curves for such high-strength
yet conducted the tests reported in
NISTNCSTAR
1-3D,
WTC
steels,
NIST
and had not
Appendix A.
Investigation
145
Chapter 6
— — ———— — ———— —— —— — —— — —
I
T
120 -
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
\
-
F =60
perimeter T=20°C
ksi
_ _ - -20 C
_ - 350°C
_ _ - -400°C
100
_--
-450°C
-
500°C
-550X
- -600
C
-650°C
20
Q
1
I
I
I
I
0.00
I
I
I
I
\
0.05
I
I
I
I
0.10
\
I
I
0.15
I
!
I
I
I
I
0.20
\
1
I
I
I
0.25
I
I
I
I
\
0.30
I
L
0.35
True Strain
09 31 ;59
-
Tuesday October
12,
2004
Figure 6-15. Example of predicted stress-strain curves for Fy = 60 ksi perimeter columns
calculated using Eq. 6-4.
Limitations of Elevated-Temperature Stress-Strain Curves
There are important limitations to the high-temperature stress-strain curves. At room temperature the
stress-strain curves are time-invariant, and, to a
good approximation
rate-invariant as well.
temperature, beginning at about 400 °C, where thermally activated dislocation motion
is
At elevated
easier, these
assumptions are no longer valid. In particular the yield strength, defined as the stress below which no
permanent deformation occurs,
Summary
of Data for All
is
very poorly defined.
WTC
Steels
Table 6-9 shows the values of Rts and Rc for
all
the different steels in
WTC
1
and
WTC 2.
Example Calculations
To
estimate the stress-strain behavior of a
F,.
= 60
ksi perimeter
column
steel
use the supporting data from
Table 6-8 and Table 6-9 with Rjs = 1.1800 and Rc = 0.8900. Figure 6-16 shows the stress-strain curves
generated using Eqs. 6-4, 6-5, and 6-6 and these parameters. Note, of course, that there
yield point in the curves generated
146
is
no
distinct
by Eq. 6-4.
NISTNCSTAR
1-3D,
WTC
Investigation
Elevated Temperature Properties
Table 6-9. Scaling parameters (Eg. 6-4) for
Static Fy
Static
WTC
all
steels.
TS
Summary
He
(ksi)
(ksi)
36.7
64.5
1.09
0.86
All 36 ksi core
box column
37.0
63.5
1.07
0.95
All 36 ksi core
WF
51.4
79.2
1.07
0.88
All 42 ksi
box column
steels
47.0
74.8
1.01
0.88
All 42 ksi
box column
steels 0.75in
42.6
70.4
0.95
0.88
All 42 ksi
box column
steels
53.8
74.4
1.00
0.98
42
49.0
71.1
0.96
0.95
44.2
66.6
0.90
0.95
53.8
74.4
1.00
0.98
49.0
71.1
0.96
0.95
47.8
71.1
0.96
0.94
53.8
74.4
1.00
0.98
53.8
74.4
1.00
0.98
WF Core Columns
42 ksi Group 3 shapes WF Core Columns
42 ksi Group 4&5 shapes WF Core Columns
45 ksi Group 1&2 shapes WF Core Columns
45 ksi Group 3 shapes WF Core Columns
45 ksi Group 4&5 shapes WF Core Columns
50 ksi Group 1&2 shapes WF Core Columns
50 ksi Group 4&5 shapes WF Core Columns
35.6
61.2
1.03
0.88
All 36 ksi perimeter
steels plates
1
2 4
(
not inner
web =
plate 3)
53.1
74.9
1.01
0.95
All 45 ksi perimeter column steels plates
1
2 4
(
not inner
web =
plate 3)
54.0
75.6
1.02
0.98
All 50 ksi perimeter
column
steels plates
1
2 4
(
not inner
web =
plate 3)
60.8
82.6
1.11
0.90
All 55 ksi perimeter
column
steels plates
1
2 4
(i.e.
not -plate 3) with
All 60 ksi penmeter column steels plates
1
2 4
(i.e.
not inner plate -plate
1
2 4
(i.e.
not inner plate —plate
ksi
87.3
1.18
0.89
3) with
69.6
90.4
1
22
0.98
steels
t
t
< 0.75m
>
<t<
1
.5
in
1.5 in
Group 1&2 shapes
thickness
62.0
Description
t
<
<
1
column
.5 in
1.25
m
All 65 ksi perimeter column steels plates
3) with
t
<
0.5 in
76.7
92.0
1.24
0.96
All 70 ksi perimeter column steels plates
1
2 4
(
not inner
web =
plate 3)
82.5
96.8
1.31
0.94
All 75 ksi perimeter column steels plates
1
2 4
(
not inner
web =
plate 3)
91.5
99.4
1.34
0.99
All 80 ksi perimeter column steels
104.8
116.0
1.57
0.98
All 85, 90, 100 ksi perimeter
38.1
59.6
1.00
0.94
Truss rounds specified as
55.3
74,1
1.00
0.96
Truss angles (A 36 or
42.6
n d
0.90
0.91
All
= 42
ksi perimeter
column
plate 3 steels
45.9
n/d
0.94
0.92
All Fy
= 45
ksi perimeter
column
plate 3 steels
51.4
n/d
1.00
0.93
All F,
= 50
ksi perimeter
56.9
n/d
1.07
0.91
All
= 55
ksi perimeter
column
plate 3 steels
62.4
n/d
1.13
0.95
All Fy
= 60
ksi perimeter
column
plate 3 steels
67.9
n/d
1.19
0.98
All Fy
= 65
ksi perimeter
column
plate 3 steels
78.9
n/d
1.31
1.00
All F,
= 70
ksi
Key:
A
A
column
plates
steels, regardless
of plate
36
242) and
column
and F, = 75
all
ksi
all
rounds specified as A242
plate 3 steels
perimeter column plate 3 steels
n/d. not determined.
NISTNCSTAR
1-3D.
WTC
Investigation
147
Chapter 6
1.2
Intersection of E(T) with elevated temperature a-c curve
Equation
for f(T)
1.0
0.8
O
o
O
(M
.0.6
0.4
Rts =
0.2
-00
1
Rc = 0.96
FJ23°C) = 59.0
ksi
I
0.0
J
1
17 35 05
-
Monday October
I
I
I
I
400
200
L.
I
600
tCc)
2004
11,
Figure 6-16. Yield point calculated from intersection of appropriate Young's modulus
and Eq. 6-4 compared with the expression for the decrease in yield strength for
structural steels in general Eq. 6-1. The correspondence is within the uncertainty of
either expression.
To
= 36
estimate the stress-strain behavior of a Fy
column
ksi core
WF steel use the supporting data from
Table 6-5 and Table 6-9 with Rjs = 1.0700 and Rc = 0.9500. Figure 6-17 shows the stress-strain curves
generated using Eqs. 6-4, 6-5, and 6-6 and the parameters above. Note that Eq. 6-4 slightly overpredicts
the already supplied
behavior
NIST
room temperature curve
has observed for other
(the solid line), but this difference is within the range
A 36 grades of steel.
— ———— — ——— ———— —— —— — —— —— — — ——
- -20°C
^""^-^
350X
—
FY=36 ksi core WF T=20°C
- -400°C
100
-|
I
I
I
of
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
'
'
'
I
'
__(0
/^''
:^
in
CO
CD
60 -
CD
.///
/
///
40
-500°C
^
//
-
CO
-450°C
__
-550°C
P/
i_
f//
600°C
20
Q
^ _--
1^
1
I
I
-
I
0.00
I
1
I
0.05
I
I
I
I
I
0.10
I
1
1
L_l
0.15
1
I
I
I
I
0.20
1
I
I
I
I
0.25
I
I
I
I
I
0.30
I
I
L
0.35
True Strain
16 10 30
-
Tuesday October
12,
2004
Figure 6-17. Example stress-strain curves for Fy = 36 ksi
using Eq. 6-4.
148
WF core columns calculated
NISTNCSTAR
1-3D,
WTC Investigation
Elevated Temperature Properties
Analysis of Creep Data
6.4.4
Background
The
literature contains \"er>'
who
Knight (1971)
first
two
Ten 50
are similar to
is
few strain-time curves for short-temi creep of structural
characterized four Austrahan steels:
US ASTM A
36
steels.
Cr and Ni
a weathering steel with
X-60
is
additions.
AS
A135,
AS AI49,
steel.
a pipeline steel conforming to
A second source
is
One
source
is
X-60, and Aus-Ten 50. The
API 5LX. Aus-
Fujimoto (1979),
who
SM50, and SM58. SS41 is similar to
TS = 530 MPa). SM58 is a Cr-Mo-V steel
characterized the creep of three Japanese structural steels: SS41,
A 36. SM50
(F,
a higher strength grade (F,
is
= 540 MPa
= 350
MPa
(51 ksi)
TS = 660 MPa). Williams-Leir (1983) reanalyzed the data m both these studies
very useful analytical model for predicting creep strain as a function of time, stress, and
(78 ksi),
and developed a
temperature for
all
seven
steels.
Fields and Fields (1988) reanalyzed Knight's creep data of AS
steel consists
of 44
range 350 °C <
tests in the
SS41
strain-time creep behavior of
of Fields and Fields, and refers
in
r<
A 149
steel.
650 °C. They also used
The data
their
set for the
model
AS A 149
to predict the
Fujimoto (1979). The analysis of this section preserves the notation
to the predicted strain
Fields and Fields (1988) represented the creep strain,
dependence
,
using these parameters as the
Sc,
"AS A 149 model."
using a function that separates the time and stress
into independent functions:
= At^a^
C are
The functions A, B, and
all
6-10
functions of temperature.
B(T)^B,+B,T
where
Bo=
5,
where Tis
in
°C and
/
in
is
6-11
-1.1
=
0.0035
minutes.
c(r) =
Q + c,r
where
CoC,=
where Tis
in
°C and c
is
in ksi.
often denoted
by the symbol
NISTNCSTAR
1-3D,
WTC
6-12
2.1
0.0064
The parameter
C is the
temperature dependent creep stress exponent,
n.
Investigation
149
Chapter 6
A(T):=\0~'
where
for
r<
500 °C
Aa^ 6.10
A^^ 0.00573
for 7 > 500 °C
Ao= 13.25
Ai
where T is
The
°C and
in
= -0.00851
strain is in percent.
analysis of the remainder of this section has
behavior of the important truss angle
steel that
develop a methodology for estimating from
A
The
A
242
goals.
made up
One
is
to characterize the actual creep
the chords of the floor trusses.
literature data the
The second
to
is
creep behavior of the other steels in the
242 Fy = 50 ksi Truss-Angle Steel
came from the upper and lower chords of truss specimen C-132, of unknown
buildings. The steel was specified to conform to ASTM A 242. Chemical analysis showed
steel tested
location in the
that,
two
columns, and core columns.
trusses, truss seats, perimeter
Results for
6-13
although
its
chemistry
is
consistent with the
usual Cr and Ni alloying elements added to
A
572
to a
contemporary
The
solid lines in Figs.
A
A 242
242 specification,
it
contains reduced levels of the
steels for corrosion resistance.
Chemically
it is
similar
steel.
6-2 through 6-5 plot the creep curves for
A
242.
The curves have been
graphically truncated at 5 percent strain. Horizontal arrows denote tests that were suspended. Vertical
arrows denote
Analysis of
The
tests that
A 242
ended with specimen
failure.
Fy = 50 ksi Truss-Angle Creep Data
analysis retains the formalism of the 1988 analysis of Fields and Fields,
which describes the creep
(not total) strain as
=
where A,
In practice
and
it
C are
all
c
6-14
functions of temperature.
was not possible
and Fields evaluated
At''(T
to evaluate both
C from plots
A and C
independently
in a single
of creep strain as a function of stress
creep
test.
at different times,
While Fields
we
evaluated
C
as the stress exponent for creep rate, ds/dt:
t
150
^C{T)
6-15
NIST NCSTAR
1-3D.
WTC Investigation
Elevated Temperature Properties
Calculating
C this way enabled the use
non-linear least squares
fit
of extra
strain rate data
to the (ds/dt, a, T) data
shown
in
from the high-temperature
A
tensile tests.
Table 6-10, using the natural logarithm of
Eq. 6-15 resulted in
C{T)^C, + C,T
where
Co=
C,
where T is
°C and a
=
3.233
0.0117
MPa. Using the entire data set simultaneously resuhs in a fit that weights the
temperatures with more data points more strongly. Figure 6-18 shows the strain rate data and the fits to
Eq. 6-16. The range of OT^ is 7.9 < C(T) < 10.8 for 400 °C < r< 650 °C.
in
is in
Table 6-10. Stress-temperature-strain rate data
used to evaluate the parameters of Eq. 6-16.
Strain
Specimen
NISTNCSTAR
1-3D.
WTC
Flow Stress
T
Rate
MPa
°C
1/s
C132TA6W-1
165
650
9.02E-04
C132TA7W-1
147.2
650
2.69E-04
C132TA6N-2
145.4
650
2.13E-04
C132TA2N-3
126
650
1.39E-04
C132TA2W-8
125
650
3.20E-04
C132TA2W-4
100
650
2.95E-05
C132TA4N-3
212
600
7.62E-04
C132TA1W-8
202
600
2.50E-04
C132TA3W-4
190
600
2.62E-04
C132TA5W-2
164
600
2.79E-05
C132TA3W-3
152
600
2.85E-05
C132TA4W-6
150
600
1.90E-05
C132TA2W-2
131
600
9.91E-06
C132TA1W-2
123
600
2.44E-06
C132TA1N-3
349
500
8.60E-04
C132TA1N-2
300
500
2.50E-04
C132TA2N-2
297
500
4.11E-05
C132TA4W-4
278
500
1.53E-05
C132TA1W-4
272
500
7.75E-06
C132TA3N-1
252
500
6.77E-06
C132TA4W-3
250
500
3.48E-06
C132TA3W-2
200
500
5.17E-07
C132TA4N-2
445.8
400
2.67E-06
C132TA7W-2
422.4
400
8.81E-07
C132TA7N-1
346.4
400
8.40E-08
Investigation
151
Chapter 6
1
1
'
1
1
1
'
1
"
0
-
A
w
.4
1
1
°C
°C
°C
°C
A
/6
0
0/V
_
0
A
/Q
/
£J
/
/A
-
-
//a
o
/A
o
1
'
1
-
650
600
500
400
0
1
'
0
-6
A
0
dcydt = A' a^'^'
C(T) = Co+C.T
1
,
1
,
1
,
2.0
,
,
2.2
-
Thursday Oclober
14.
,
,
,
2.4
logio(cj),
19-21 36
1
1
\
,
2.6
MPa
2004
Figure 6-18. Prediction of the function C(T), Eq. 6-16, from strain rate data for
truss steels.
Unlike the data for the
AS A 149
steel, the
A 242
time-hardening exponent, B, does not increase linearly with
temperature, so the representation of Eq. 6-1
1
is
inappropriate.
To ensure
accurate representation of the
creep behavior near and outside the boundaries of the experimental conditions requires choosing a
function for
at
B
that
does not diverge or produce negative values of B
temperatures less than about 400 °C
is
at
insignificant, so the specifics
temperatures will not affect the measurable
strain.
The chosen function
low temperature.
of the behavior
for B(T) has a
Of course,
at
creep
low
minimum
value for
low temperature:
B{T) = B,,+BJ^'
where
5o=
5,=
82=
where
7" is
in
°C and
/
is
in seconds.
3.5531x10""
3.6975
The parameters have no physical
behavior of B(T) over the temperature range where creep
152
6-17
0.3982
is
significance. Figure
6-19 shows
the
important.
NISTNCSTAR
1-3D.
WTC
Investigation
—
r
Elevated Temperature Properties
2.0
T
—— — — —
I
\
I
I
I
'
'
'
'
'
'
'
I
I
'
'
'
I
I
o
B(T) = Bo+B,t"^
O
1.5
o
m
1.0
0.5
0.0
'
'
350
'
'
'
I
'
'
'
I
400
I
I
I
450
I
I
I
I
I
-
I
I
I
I
I
I
I
600
I
I
I
I
650
I
I
700
°C
S with
6-14 could be represented by
prefactor, A, in Eq.
I
I
Thursday Ociober 14 2004
Figure 6-19. Variation of the parameter
The
I
550
T,
19:22:55
I
500
A{T) =
^,,=
exp(A^^
temperature. Solid line
is
Eq. 6-17.
a quadratic polynomial.
+ A^T +
y^,^'
)
where
-55.4504
6-18
9.47600x10"^
Ai=
A2= -3.52064x10-'
where T
is
in
°C and
strain is in natural units.
Evaluating the parameters of Eqs. 6-16, 6-17, and 6-18 required several steps. The
evaluate d, and C/ of Eq. 6-16 using the data of Table 6-10. After that, Eq. 6-14
first
was
fit
step
to
was
to
each
individual creep curve, using the appropriate values of C(T) and a. This resulted in a set of 20 individual
(Ai.Ti)
and
(B,,!,) pairs,
linear regression in
Comparison
from which the values of the parameters
in
Eqs. 6-17 and 6-18 were evaluated by
the case of Eq. 6-18 and non-linear, least-squares
of the
A
fitting in
Eq. 6-17.
242 Model with Actual Creep Curves.
Figures 6-2 through 6-5 also plot the model creep curves developed using the methodology of the
previous sections. In general the agreement
of WilHams-Leir (1983)
to his
own
data.
It is
to
is
acceptable and
is
of the same magnitude as the model
Knight's (1971) and Fujimoto's (1979) data, as well as Fujimoto's
important to realize that the model curves in the figures are not
creep cur\ es. but are rather the prediction of the model from the global
fit,
fits to
(
fits
1979)
fits
the individual
using the methodology of the
previous section.
NISTNCSTAR
1-3D.
WTC
Investigation
153
Chapter 6
Estimating Creep Properties of
The general goal of the second
WTC
properties of the other
creep data from the
NIST
WTC
Steels.
part of this analysis
is to
develop a methodology to estimate the creep
using either literature-generated models or models generated from
steels
laboratory.
Previous analysis has indicated that both the yield (F,) and tensile (TS) strengths follow a universal curve
with temperature, when normalized to their room temperature values. This behavior suggests that
it is
possible to use the creep data developed for one steel to predict the creep behavior of another steel by
appropriate scaling by one of these parameters.
total elongation to failure are frequently the
AS A 149 model
to predict creep
to the applied stress,
untested
steel.
cr,„
A
second advantage of this approach
only steel parameters reported in the
behavior of an untested
steel,
To use
literature.
AS A 149
and
TS,
F,.,
Eq. 6-14 uses a corrected stress,
times the ratio, R^, of the yield or tensile strength of the
the
equal
ac,
to that
of the
6-19
^^a^a
^
So, for an untested steel that
is
^ um
stronger than the reference
AS A 149
predict the creep behavior will be less than the actual applied stress
is
that
For example,
(^c
which
is
even simpler,
is
=
to set
1
,
which
is
steel, the
(Ra<
!)•
equivalent to assuming that
corrected stress used to
An
alternative approach,
all steels
have identical creep
properties.
The
best of the three possible scaling ratios, R^. (Fy
,
TS,
and no scaling^ was evaluated subjectively.
Williams-Leir's (1983) analytical expressions for creep strain as a function of time for the four Australian
steels
of Knight (1971) and three Japanese steels of Fujimoto (1979) facihtated
Although Williams-Leir used an expression for creep
that differs
from Eq. 6-14,
his fits
expression removed the need to
and 650 °C) the
to that predicted
steel.
ratio,
strain as a function
produced creep curves of similar
re-fit all
seven data
R, of the creep strain
at
3600
this intercomparison.
of time,
fidelity.
and temperature
stress,
Using
his analytical
At four temperatures (500 °C, 550 °C, 600 °C,
predicted by the scaling ratio approach of Eq. 6-19
sets.
s
from the Williams-Leir analytical expression was plotted as a function of stress for each
The range of stresses
plotted
was constrained
to
produce
strains in the range
5xlO"^<£<0.25.
Strains outside this range are irrelevant for creep modeling. This procedure produced 28 plots (seven
steels at four temperatures)
where each
plot has seven plotted ratio lines. Perfect prediction
another's creep strain would produce a horizontal ratio line with
ratio, Ra-, the quality
R=
I.
For each
steel at
by one
steel
of
each scaling
of the prediction was ranked by visually judging the distance from the distance of the
predicted line from unity.
The
results
of the ranking established the tensile strength
of the cases, tensile-strength
ratio scaling
cases, tensile-strength ratio scaling
ratio scaling as the best approach. In
and no scaling produced similar
produced the best
results,
and
in the
results. In
about
1/3
about 1/2
of the
remaining 1/6 of the cases, no
scaling produced the best results. Thus, in about 5/6 of the cases, scaling the creep behavior by the
room
temperature tensile strength ratio produced creep curves that most closely approximated the measured
curves. Scaling by the yield strength ratio almost always produced the worst predicted strains. Frequently
it
produced estimates
154
that differed
from the actual data by more than a factor of ten. The only times
NISTNCSTAR
1-3D.
WTC
that
Investigation
Elevated Temperature Properties
scaling
by
the yield strength ratio produced acceptable results
identical to the scaling ratio
produced by the
was when
the scaling ratio
was
fortuitously
tensile strengths.
Values for Creep Parameters
To
estimate the creep strain as a function of time for
parameters
To
6-14 with the
all
truss bulb angles use Eq.
all
other steels use Eq. 6-14 with the parameters of
6-16, 6-17 and 6-18.
in Eqs.
estimate the creep strain as a function of time for
Eqs. 6-11. 6-12 and 6-13. Scale the applied stress by the ratio of the tensile strength of the
steel {TSai49
=
AS A149
70.6 ksi) to that of the steel in question using Eq. 6-19.
Values for Bolts
6.4.5
Background
NIST
did not test any bolts at elevated temperatures. Given the lack of experimental data,
to predict the bolt properties using the
A 325
Because
and
A 490 bolts
temperamre property data
methodology of Section
are not intended to be used
The
exist in the literature.
it
is
necessary
6.4.1.
above room temperature, very
little
elevated
limited strength data for bolts plotted in Fig. 6-20
come from tensile specimens of bolt steels rather than the bolts themselves, made to British, Japanese and
Chinese boh standards. In general, these bolt standards are similar to the corresponding ASTM standards,
however. Table 6-1
1,
and the
rest
of this section, summarize the literature data used to construct
Fig. 6-20.
Ta ble 6-1 1
TS
Reference
Steel
Sources of
.
bolt data.
Fy (Spec)
(spec)
MPa
MPa
Steel
Chemistry
Kirby
Grade 8.8
785-981
628
C-Mn-Cr-Mo-B-Ti
Li
20MnTiB
1040-1240
940
C-Mn-Ti-B
Bol-Ten
981-1177
883
unknown
827
634
C-Mn-B
Sakumoto
A
325
BS 3692 from two manufacturers. These bolts
and chemistry. He conducted tensile strength tests at
Kirby (1995) studied British Grade 8.8 bolts supplied
are very similar to
A
325 bolts
in
both strength
to
temperatures up to 800 °C on actual bolts, as well as on the steels used to fabricate the bolts. In the bolt
tests, the
mode of tensile
failure
depended on choice of nut. Even though the nuts were made
specifications, in one set the bolt failed
by ductile necking
in the thread,
while
to identical
in the other set, the bolt
threads stripped out.
Li et
al.
(2003) tested bolt steel identified as "20MnTiB," which apparently corresponds to a Chinese bolt
specification. In terms of strength, these bolts are similar to
tested tensile specimens fabricated
NISTNCSTAR
1-3D.
WTC
from
Investigation
A 490 bolts, rather than to A
bolt steel at temperatures
up
to
325
bolts.
They
700 °C.
155
Chapter 6
Sakumoto (1993) reported the properties of ordinary and "fire-resistant" (FR) bolts made
1981. Both were identified as "BOL-TENl ION" steel. This bolt steel is stronger than the
to
JSS
A 325
11-09-
bolt
steels.
T
1.0
0.8
E
0.6
cn
t:
cn
O
5
«
6
0.4
TS Sakumoto B0LTEN1 10N-FR
TS Sakurnoto BOLTENl 10N
20yr!TIB
TS:
Li
5%
Stress: Kirby Set
Recommend
Recommend
A Gr
8,8 V-B
value for structural steel
value for bolt steel
0.2
0.0
_i
i_
I
_I
I
L
200
T,
09:41 45
-
J
400
I
l_
600
°C
Friday August 27, 2004
Figure 6-20. Comparison of high-temperature yield, Fy, and tensile, TS, strength for bolt
steels and the expression for structural steels in general, Eqs. 6-1 and 6-2. Solid
symbols are bolt steels. Open symbols are "fire-resistant" bolt steels. Expression for bolt
steels is the dashed line.
Figure 6-20 shows the ratio of yield
the various bolt steels.
156
The
solid line
(F,.)
is
and
tensile (TS) strengths to their
room-temperature values for
the expression for structural steels in general.
NISTNCSTAR
The
1-3D,
bolt steels,
WTC
Investigation
Elevated Temperature Properties
however, with the exception of the
FR BOL-TEN
steel, lose strength
more
rapidly than do the structural
steels.
Results
Because the bolt
is
steels lose strength
more
rapidly with increasing temperature than do structural steels,
necessary to use a different set of parameters for Eqs. 6-1 and 6-2 to describe the strength
it
loss.
Table 6-12 shows the values for the parameters for yield (Eq. 6-1) and tensile (Eq. 6-2) strength
reduction.
The dashed curve
in Fig.
6-20 corresponds
to the expression for the strength reduction
of bolt
steels.
Table 6-12. Values for the parameters
and 6-2)
\'ie\d
the strength reduction equations (Eqs. 6-1
In
for
use with bolts.
Strength Reduction Parameters
Tensile Strength Reduction Parameters
(Eq.6^1)
=
(Eq. 6-2)
0.07
A2=
0.09
w, =
6.8221
/7,
=
5.0237
W2=
1.0000
«2
=
5.0245
.42
s,
= 491.0
S2= 810.0
t2
= 478.0
= 585.0
Values for Bolt Strength Parameters
The
yield and tensile strength of bolt steels degrade
structural steels.
more rapidly with temperature than do ordinary
The reductions can be represented by Eqs. 6-1 and 6-2, using
the parameters in
Table 6-12.
Limitations
Most of the data
in Fig.
6-20 come from
(1995) provides data from actual bolt
tests
tests.
on bolt
No
increasing temperature for structural steel bolts.
cannot be directly applied to predict bolt
relatively
little
steels, rather
failure,
The
ductility
measured
in tensile tests
because bolts generally
fail in
of bolt
steels
the threads. There
is
data on the expected (versus specified) strength of bolts. Fisher (1963) and Salih (1993)
found very different values for the average strength of A 325
strength
than on bolts themselves. Only Kirby
information exists on changes in failure mechanisms with
was much higher than
bolts,
though as expected, the average
the specified strength.
In the absence of data to the contrary, the reduction in bolt strength with temperature can be
modeled with
Eq. 6-2 using the parameters of Table 6-12.
6.5
The
SUMMARY
yield and tensile strength of structural steel follow a master curve
when normalized
to their
room
temperature values, Eqs. 6-1 and 6-2.
NISTNCSTAR
1-3D.
WTC
Investigation
157
Chapter 6
The shapes of stress-strain curves
at
elevated temperature can be modeled based on the room-temperature
shape of a model curve scaled by the ratio of their tensile strengths. The model accounts for the decrease
in
work-hardening with increasing temperature, Eqs. 6-4, 6-5 and 6-6. The behavior of steels with
= 36
Fy > 36
F,.
ksi
is
ksi is
based on the behavior of A 36
based on the behavior of F,
Creep behavior can be modeled using
steel
= 50
from the
literature.
ksi steel recovered
The behavior of steels with
from the
a function that breaks the time
and
WTC
stress
floor trusses.
dependence of the creep
= 36 ksi is based on a literature
= 36 ksi steels, it is necessary to
strain into separate functions, Eq. 6-14. The behavior of steels with Fy
analysis of Australian
AS A 149
scale the applied stress
by the
steel.
ratios
behavior of steels with Fy > 36 ksi
is
To apply
this
model
to other F,
of their room-temperature tensile strengths, Eq. 6-19. The creep
based on results of tests on Fy = 50 ksi
steel
recovered from the
WTC floor trusses.
Bolt properties are estimated from literature data.
REFERENCES
6.6
The
Baird, J.D. 1971.
metals.
Review
Chijiiwa, R. et
Nippon
al.
effects
of strain-ageing due to
interstitial solutes
on the mechanical properties of
149. Melall. Rev. 16 1-18.
1993.
Development and
Practical Application of Fire-Resistant Steel for Buildings.
Steel Technical Report 58. 47-55.
Fields. B.A.
and
R.J. Fields. 1989. Elevated
Temperature Deformation of Structural
Steel.
NISTIR
88-
3899.
Fujimoto,
M.
e/ al.
1979. Primary Creep of Structural Steel
Laboratory of Engineering Materials, Tokyo
Goda,
at
High Temperatures. Report of the Research
of Technology.
1964. Present Status of Weldable High Strength Steel.
S. et al.
Seitetsu
Institute
Kenkyu) 248 5086-5163
Number 4.
p.p. 1-27.
Yawata Technical Report (Yawata
(in Japanese).
Gowda, B.C. 1978. Tensile Properties of SA 516, Grade 55 Steel in the Temperature Range of 25°-927°C
and Strain Rate Range of lO"'' to 10"' sec"'. In, Characterization of Materials for Sen'ice at Elevated
Temperatures. Smith, G.V. ed. American Society of Mechanical Engineers. New York, "presented at
the 1978
ASME/CSME Montreal
Pressure Vessel
& Piping Conference, Montreal,
Quebec, Canada,
June 25-29 1978." 145-158
Harmathy T.Z. and W.W. Stanzak. 1970. Elevated-Temperature Tensile and Creep Properties of Some
Structural
and Prestressing
Steels, in Fire Test
Performance,
ASTM STP 464, American
Society for
Testing and Materials. 186-208.
Holt, J.M. 1963. Short-time Elevated-Temperature Tensile Properties of
A Constructional Alloy
reported in
158
(USS
Steels.
USS
"T-l" and
USS
"T-l" Type
United States Steel Report (57.019-901)(3)(a-AS-EA-2). Also
1972).
NISTNCSTAR
1-3D.
WTC Investigation
Elevated Temperature Properties
Holt. J.M. 1964. Short-Time Elevated-Temperature Tensile Properties of USS
High-strength Low-alloy
steels,
USS Man-Ten (A
Cor-Ten and
440) High-strength Steel and
USS
ASTM A
Tri-Ten
36
Steel.
United States Steel Corp. Report 57.19-901(1). Monroeville, Pa, Applied Research Laboratory.
Oct. 16
(much of this data appear
Jerath. V.. K.J.
in
(USS 1972)
Cole and C.L Smith. 1980. Elevated Temperature Tensile Properties of Structural Steels
Manufactured by the British Steel Corporation. Unpublished Report # T/RS/1I89/1 1/80/C. British
Steel Corporation, Teeside Laboratories. July 24.
Kirby, B. R. 1995. The Behaviour of High-strength Grade 8.8 Bolts in Fire.
J.
Construct. Steel. Research.
33 3-38.
M.G. Lay. 1971. Short Term Creep Data on Four Staictural Steels. Report
Melbourne Research Laboratories. The Broken Hill Proprietary Company. Clayton,
Knight. D.. D.H. Skinner and
MRL
6/3
Victoria, Australia. Januar}'.
Li.
Guo-Qiang. Shou-Chao
Jian,
Ying-Zhi Yin. Kai Chen and Ming-Fei
on the Properties of Constructional Steel
at
Elevated Temperatures.
Li.
2003. Experimental Studies
J. Struct.
Eng. 129[12] 1717-
1721.
Manjoine, M.J. 1944. Influence of Rate of Strain and Temperature on Yield Stresses of Mild
ASME.
Steel. Trans.
A-211-A-218.
66.
Melloy, G.F. and J.D.Dennison. 1963. Short-time Elevated-Temperature Properties of A7, A440, A441,
V50, and V65 Grades. Internal memorandum
to J.W.
Frame. Report 1900-4g. Bethlehem
Steel.
April 17.
Rumpf
John L. and John
W.
Fisher. 1963. Calibration of A325 Bolts.
J. Struct.
Div.
ASCE. 89[ST6].
215-234.
Sakumoto, Y., K. Keira,
Struct.
Eng.
Salih, Nazir,
1
F.
19[1 1]
John Smith,
A325 High-Strength
Furuyama and
T. Ave. 1993. Tests of Fire-Resistant Bolts
and
Joints. J.
3131-3150.
el al.
Bolts.
1992
An
J. Test.
Experimental Appraisal of the Load-Deformation Properties of
Eval. 20[6] 440-448.
(USS) 1972. Steels for Elevated Temperature
Sen'ice. United States Steel Corporation. Pittsburgh, Pa.
Williams-Leir, G. 1983. Creep of Structural Steel in Fire: Analytical Expressions. Fire
and Materials.
7[2] 73-78.
NISTNCSTAR
1-3D.
WTC
Investigation
159
Chapter 6
This page intentionally
160
left
blank.
NIST NCSTAR
1-3D,
WTC
Investigation
Chapter 7
Summary AND
Findings
SUMMARY
7.1
The experimentally determined mechanical property
data of World Trade Center
(WTC)
steels
can be
divided into five groups:
•
Elastic properties as a function of temperature
column
•
Room
were determined from recovered perimeter
steels.
temperature yield and tensile strength, and total elongation were determined from
specimens of recovered
steel
strain curves are reported for
of all grades from the
29 different
fire
•
stress-
with 12 yield strength levels. Load-
steels
A
displacement curves were measured from recovered
strengths of several different perimeter
and impact zones. Complete
column and
325
truss
bolts. Stress-strain
curves and
weld geometries are reported.
Yield and tensile strengths and total elongations of selected perimeter and core column steels
were determined as
a function strain rate at rates
complete stress-strain curves for these
up
to
400
s"'.
Strain-rate sensitivities
steels are reported. Several steels
and
were characterized
at
higher rates using Kolsky bar testing.
•
Impact properties as a function of temperature were determined using Charpy
perimeter column, core column, truss, and truss-seat
•
tests for selected
steels.
Elevated-temperature yield and tensile strength, and total elongation were determined on
selected specimens of recovered perimeter and core column, truss, and truss seat steels.
Complete stress-stram
characterized.
cur\'es for temperatures
Creep deformation as
a function
up
to
650 °C are reported
of temperature and
stress
for all steels
was determined
for
the truss chord steel.
By combmmg
these
recovered mill
test reports.
measured properties with
historical averages
from the
literature,
and
in
some cases
National Institute of Standards and Technology (NIST) developed expressions
and values for properties of the important grades of steel:
modulus and Poisson's
•
Elastic
•
Model room-temperature
ratio as a function
of temperature,
stress-strain curves for all
WTC steel grades from the fire and
impact zones corrected for dynamic effects,
•
Room-temperature load-displacement curves for
•
Yield and tensile strength for weld metals,
•
Stram
NISTNCSTAR
rate sensitivity
1-3D.
WTC
of
steel for
Investigation
A
325
bolts,
high strain rate properties.
161
Chapter 7
•
Elevated-temperature yield and tensile strength and complete stress-strain behavior for
all
WTC steel grades from the fire and impact zones,
•
Elevated-temperature load-displacement curves for
•
Creep response
for all
A
325
bolts,
WTC steel grades from the fire and impact zones.
FINDINGS
7.2
This report makes a number of findings concerning the steel used in the
•
Steels, bolts,
to
•
and welds generally have properties
WTC:
that indicate that they
met the
specifications
which they were supplied.
Infrequently, the measured yield strengths are lower than called for
specifications.
The measured yield
strengths of 2 of 24 perimeter
by
the appropriate
column samples were lower
than the specified minimum. This probably occurred from a combination of the natural
variability of steel properties,
The measured
protocols.
specified
minimum.
and small differences between the mill and NIST
yield strengths of 2 of 8 core
In this case, mechanical
damage
testing
column samples were lower than
that occurred in the collapse
the
and
subsequent recovery removed the expected yield point behavior leading to the low values.
•
The average measured
specified
minimum
yield strength of the steels
values by about
1
from the perimeter columns exceeds the
0 percent, which
is
consistent with historically expected
values for steel plates.
•
The measured
yield strengths of the F,
- 36
ksi wide-flange core
expected from historical measurements of other structural
•
The
strain rate sensitivities
The impact
from the
column
•
162
from the
WTC construction era.
properties of the steels, evaluated
by Charpy
testing, are similar to structural steels
WTC era. The ductile-to-brittle transition temperatures of the perimeter and core
steels are at or
The behavior of the
that
steels.
of the yield and tensile strengths of perimeter and core columns are
similar to other structural steels
•
columns are lower than
below room temperature.
yield and tensile strengths of
of other structural
steels
from the
WTC steels with temperature is similar to
WTC construction era.
NIST NCSTAR
1-3D,
WTC Investigation
Appendix A
Data Tables and Supplemental Figures
A.1
DATA tables FOR ROOM-TEMPERATURE MECHANICAL PROPERTIES
The following
NISTNCSTAR
tables
1-3D.
summarize the room-temperature mechanical
WTC
Investigation
properties.
163
Appendix A
CM
_j _J _1
_l
LL LL LL
Cl
E
I—
CO
ro
CM CM
hp _J _J
LLu!.uI.U.-j-j-jH-H-j-jHHK^^h"'T'T
LlLLLlLlLlU-LlU-LLU-
m CD CD
oOO
-Ml
CD CQ CQ CQ
ch CO ro
LU LU LU
CT3
CNj
•^rvicO'^CMcNiro'^rO'^j-CMco^CNiro
OOO
CO
to
<<< oo
CD CD CD CD
-Ml
•Ml
IW-
i
CM CM CM CM
O CJ O o
00 00 o6 CO
ZZzZ
OOOOOOOOOOOOOOOOOOOOOOOOO
C/}
zzz zzz
zzzz
T3
'd-CMrorO'^OOCT)
CD CD CD NCM CM •<-
O
CD
O
CM
CNJ
oo
CD CD
CO
CD in in CD CD CD CD
O
(U
0)
CM CM CM CM
CO CM CD LD
>-
r\j
r-oO'<-o-<-or-roaiLno
ro CM
r-cMrorororocMOcncno
CD
(35
CD
ro CD CD CD
o
o
in siCO
CO CD
00
ro CO CD sT in
in in in CM CM CM in
CD CO
CD in CO
in CD CD
J
(A
c
£
—
.
O
o
c/3
0)
0
a
c c c c c c c C c c c c r c c c c c c c c c c c C c c c c c
C c c
m
-p
5- CM CM CM
ro CD CO
ro CM CO
O
CM
C35
m
CM
o
in
O
<^
iri
CM CM CM CM CM CM
CM CM CM
CM CM CM CM CM CM CM
r\i
O
oO
(35 CD in in ro OO
en CD in
CD rr- CD
ro CO ro CO ro CM CM CM CM CM ro ro CM CM CM CO CO CO CM CM CM CO ro
O
o
o
05
O
05 ro
CM CM ro CO
o
^
in CJ5
00 CD CM h~ ro CD
CM
00
in
CO
00
CD
00 CM 00
r-1^ CD
in in ro
in in CT5 CD oo 00 00
CD
CD in rCD in in
in in in co CD CD CD CD CO in in in
in in in
in
in (D CD CD CD in in in in in in
sl-
in in
-C
v_
CM CM CM CM CM CM CM
o
m
O
m
O
m
m
in CD
in
ro CD in (35 h-- r- C35 CM CD C35 CM
00
CM ro CD CO
CD in
C3) r~CM
ro ro
00
CD in
CD
in CD CD iri CD
C35 00 00 CM (>j ro CM CM ro ro
1^ 00 00 00 CO oo CO 00 00 CO 00 CO CO CO 00 00 CO 00 00 00 00 OO (35 C35 C75 CD 05 05 oo 00 r-
d
J
(/)
3
(A
V
^ - Boo
o
in in LO CO
CO C35 00 05
00
h-in in in CM CM CM
CO in in CD CD CM CM CM CM CD
ro
ro CO ro CO ro
r--
in
in CM CM CD CD ro CD ro ro
ro CM CM CM CM CM CO CJ5 CT5 CD 05 CO CM
ro
CD
ro 5rvj
C/3
(0
0)
0)
7
c
- Bon
o >
oo
o
00
in CO CM
(35
ro CO CM
CD
in in in in in CD
rsj
CD
ss5
oO
o
O
O
o
in
in CD
in
05 CD 00 CD
ro CD ro CM
CD
CD CM CM in
CM
CD
05 CD CD
CM
CO ro
CM
CD CM
CD ^' iri CD
CD in in in
in CD CD CD CD CD CD CD CD CD
CD CD CD CD
h-
(/)
0)
-05
>-
l-
3
.E Qo >•
in CL in
CD Q. CL CL CL CL D- CL 05 CJ5 Q.
CL D. D. CL CL CL D. CL CL D. D. CL CL CL CL CL
>- iri iri
>CD 00
C35
CD
CD cb CO
in
CD in
in in in
CD CD
CD
>> > >> >
ZzzZZ ZZ
>
z
> > > > > > > > > > > > > > >
ZZZZZZZZz ZZ Z Z ZZZ
>
Z
rororoinr^'^cO'rrrMroooroooooinoooinooocDinoincoi^cMinoooocD
a
cDCDCD'^'^in'Tj-cDin-"—
CMC350T-mroLnrorot--^-r-c35COLn'^cO"^cDr^oO"^ro
c35cy5CJ5r-r~-CDC3)C75CJ5rorocMrororororororororororororororocorococorororo
^.-OOCMCMCMCM'^'^'^-<--r-^'--^'^'^-^'^'-'<-'^'^-^-r^'^-^'^'^'^'^-^'^
< - 0CDCDCDCDCDCjC3CDC30Cj0CDCDC3CD<30C5CDCDCDC3(DCDC^
E
(U
<-»
E
o
o
t
o
CD
(u
OJ
Q.
Q. >,
CO
-
.£ _|
E CD
o
C
CM^vl^^l^^OgKl^^l^g0^10g^^lO^J^vIKlKl^vl^^J^^I^^J^^^O^I
llllll-^-Sllllllllllllllllllllllllllllllllllllllllulllllllllllllllll
LL LL
Q.
I— I—
I
ooooooooooooooooooooooooooooooo
Q^Q^Q^CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCM
<
O
£1
_1_I_J|-I-I-HI--J-I
-1-11CO
c
5
o
c
c
D
CQ
CD
p
id
^
o
o
164
c
g
o
^cc
O
ro ro ro ro ro ro
05 05 05 05
OO
oooooooooooooooooo oo oo Oo OO
o NOO
Oo o
05 CD CD CD
rsi
CD C5 C5 C5
CM ro ro ro ro
f-^
C35
05 05
(35
1
CM CM CM CM
ininininLninininLnininininininLnininininin^
CMCMCMCMCMCMCMCMCMCMCMCMCMCM
D
--^
0
—
Q.
CO
c
g
o
^cc
3
inininoooooo
•^^-^inLninLninLn
ininininLninininininininLnLnLn ininLninininoooo
LnininininLnLnininLnmLnLnLnin ininLnininincDiDcDcD
gj
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
<- CM CO
CO
1— 1—
CNJ
CNI
1
1
1
§
o oo o O
m
m m m m
ro ro CO
n
CJ O o O a
CO CO CO QO 00
Z Z z Z z
Q.
E
CD
CO
CM
CM CO
LL LL U_ U.
_l _i 01 01
ul
Li.
LL.
CM CM
CM
_l _l
H
CM
H-
CM
1-
L3
CM
_J _J
LL LL
CO CO CO CO CO
RF-
RF-
-IW
-IW
-IW
-IW
CJ)
<y>
O a: 01 01 D Q D Q Q Q < < UJ LU LU < < IW i q: 01
i i
i
Hi
m
m m CDCDDDCQCDCDCDCQ
CM
00 CO
ro
00 ro
CO ro CO CO
CJ o CJ o CJ O O O
O
O
o o o CJ
CJ o CJ o CJ CJ O CJ
CD CD
(D CD
CD CD CD CD
CO CO CO CO CO CO CO 00 00
00 00 CO 00 CO CO CO CO
CD
a>
CD
ZZzZZZZ ZZ zzzzzzzzouZZ z Zzzzzz
IW-
Ljj
CNJ
C3M1
cy>
C3>
<y>
C3>
C32
CJ>
o
CM
(D
CD
OCJiOOOO'^'-^T-T--^,-^
CD
CD
0)
3
C
3C
o
"3;
CO
oO
(32 CD
CD (D CD
(35
d
00 CO
CD
^
r--
CD
<o
o o CO
oo
i^i^oocrit^'rfococaocNjt^rorvitfj'^LnaDcncTi'^Lncoroinooooco-^T-cTio
O
J2
c c c c
o
©
CD
cri
c c c c c c c
CM CM CM CM CM CM CM CM CM CM CM
UJ
O
c c
c c
j3
(TvJ
O
o
CO
CO
CO
CM
LO
LO
CO CD CD CD CD CO CD CD
CO CO
ro LO CD 00
CM CM CM CM CM CM CO CO CO CM CM CO CO CM CM CO ro CM CM CM CM ro CO CO CO CM CO
CM CM CM CM CM CM
to
c
E
m
_3
O
u
O
OO
CO
CM
CM
CO CO
CD CD CD CD
00
E
(£>
O
CO CD CD CM
CD
CD
CM 00 rCM
CM 00
OO LO 00 h- LO
CM
LO LO LO CM CO
CM CO
ro ro ro 'aCD CD CD CD CD CD CD CD CD X) CO in LO lO LO CO CD CD LO LO CD CD CD CD CD CD CD CD CD
O
o
O
o
o
(^J
CM CD
O
^
O
^
^
LO CD
CD
ro
CM CM
CD
O) CD
CO (3>
CO CD CO
CO
LO CD
CO
ro
CD
CM
LO
CD
00
CM CD CD CD CD
00 00 CD 05 CD
CD CD CD
CO ro CO
CO
CM CM CM CO CO
CD CO CD CD CO 00 CO 00 OO 00
r~-CO CO CD CD 00 00 CD CD CD CD CD CD CD CD CD
r-
CO
O
CO O
^
O
in
'hi
o>
a
o
>*-
o > ^
CM
CD
CO
00 00
CO
CD
'd-
CM
CD CM
CD rCO
r--
oo
O
CD
r-CD CD co
ro ro ro
LO LO OO CO CD CD CM
LO
CO
ro
LO CD CD
lO LO LO
Oo
3
(/)
(A
a>
o
O
o
CO CM
00
CM
CM
CO
00 CSJ LO 00
CO CM 00 LO CM
CO ro sCM CD
CO CD CD d) CO
00 oi 00
LO 0C3 00
CD CD 00 00
CD CO CO
CD in CD CD CD CD CD
CD CD CD CD CD CD lO LO LO LO LO LO CD CD CD CD r-
o > ^
LO CD
CM
ro LO LO CD
CD CD CD CD CD
CO
CL D. CL D. D. D. D. n.
'w
o
c
> > > > > > > > o
z Z z Z Z Z Z Z
>-
O CL CL D. CL CL CL CL CL
CL CL Q. CL CL o CL CL CL CL CL CL
> > > > > > >• > > > > > > > > > > >
OoO>
z z Z Z z z z z Z Z Z Z Z Z Z Z z z z
CO
CO CO
CO rCD CD 1-^
0>
LO LO
O
o CO o
CO LO CD CO CO
CM
o
O o CD
o CD
O CO ro CO CO CO CO CO CO CO
oOOOOOOOoOOO OOOOoOOOOOo oOoOOO ooooo
ooorMCMCMOOOCDLO-^'-r-~r--cDcor--.cooo
O'-CM'S-CD^LOCDCDCOr-LOLOLnCD-^OO-'-OCO
a>
a
E
a>
E
o
o
LO CO
COrOrOrOrOC0CDCDCDCDCDCDCDCDCDC0C0C0C0C0CNJ^~-
3
T3
C
CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM
CD
O
flj
(U
Q. 9:
CO
E
°
O
(/)
g
CD
CD
CD
CD
CD
CD
to
CD
LL LL LL LL LL LL LL
CD
CD
CD
CD
CD
CD
LL LL LL LL LL LL
CO
CD
CD
CD
CD
(J (J
OO
CMCMCMCMCMCMCMCMCM
CD
TO ro ro ro
LLLLLLLLLL.iliiii.iiiiii
LL
LJ.
L. LL
LL LL LL LL
E
O
T-
CD
oooooooooooooooooooooooooooooooooo
o CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCM
o
OJ
I
<
52
X
t/5
a>
o
"
_©
CD
C
c
2
u
<D
—I
0O0O0OOOOOOOOOO0OOOOOOO'<-'^CMCMCMCMCMCMCMCMCM
oooooooooooooooooooooooooooooooooo
(U
CDCDCDCDCDCDCDCDCDCDCDCDCDCDaJCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCpCD
CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCVJCDCDOOOOOOCOOOOOOOM
9
05
—
CD
^Tj'^Tt-^-^TfT^TtT3-T;tT^T;f^fTf5^T3-Tf'<;f'<d-'^-<a--^COCO^'^'^-^'*'*-^
^
^
O
u
OOOOOOOOOOOOOOOOOOOOOOOOOinLOLOlOLOLOLOLOLO
CDCDCDCOCOCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDtDCDCDCDCDCDCDCDCDCDCD
Q.
CO
0
NIST NCSTAR
1-3D,
WTC
Investigation
165
Appendix A
_i _i
CL
E
CD
H
•<- CM
CM _l _i
I-
^ CM
HH
t-CM'-CMt-CM-^CM
_J_ll-;-|-;-_J_Jl-;-l7
1
CMCMCMCMCMCMCMCM
_l _l _l _l
LL LL LL Li.
i
->
a: cc DC cr
1
[
r
1
[
cc
^
-l-L-H-LLLLLLLLrn
ig^ m £
LLLLLLLL''''!
ro ro ro
''^^^^COCDCOCOf^
UO
2225rsirMCMCM-^'<-'-'-Lj
CMCMCMCMOOOOOOOOcQ CO CO CD
S'S'S'^cMCMCMrMLnLninino O O O
1
CO CO CO CO 00 CO
CM CM CM CM ro ro
oUOoOO
CD CD CD CD 4 4
o O o o W CO
T-'^T-T-CMCNCMCMCMCMCMCM'-
COCOCOCOOOOOOOOOS
i i i
D
p
in
CD
0)
- O
CM
rO'^roinCMroO''-'-
00 ro
o
oO
CT) 00
CM CM CM
00
a>
o
oOO
LD
CD
CO
CD ID CD CD 'd-
1---
si-
o
CD Ln
CD CD
o
*
CD N- LO
C» rsi CM
CO LD LO CD CD
'l- un CD
CD
ro
q:
c c c c C C C c c c c C c C c c c c c C c C c c c c
CM
CM CM
LU
CNJ
CM CM CM CM CM
rM CM CM CM CM CM
r\i
CM CM CM CM CM CM
0'<-ocMOoO'-CMr-'-cDO'<-i--'a-'*oor-'>--^h-(j)OoocM'^
rOrOCMCMrOCMCMCM-<-CMCMrOCM'<-CMCM'<-'<-CMrM'<-f-CM'»-'^'<-
0
77,
£CD
c
OJ
«
I—
o
O
o
*
o
NCM en ro Ln CM CO
CD CD r-- ro Ln en CO CM
ro ro ro CM
ro ro
1^
00 00 00 00
CD CD CD CD CD CD CD CD CD CD CD CD
LD in CD CM in
ro CM
CD CD CD CD
o
* o oo
CO
^ -co
^
00 TCM
00
CM
cr>
a3
o
r--
CM >CM CD
m
o
Ln
CM "3- en Ln CD
00 ro CM 00
CO Ln
ro 00
CO
CD CD (D
ro ro CM CM CM
CM ro ro ro ro CM ro
CM CM CM
CD CD CD CD CD CD
CD CD <T> CD
OOOOO
oo
in CD
in
in CD
00
CD
ro
ro in CD in in
in
in
CD
in
CM ro ro
LD Ln in in in in Ln in in in in in in CD CD CD CD
r~-
C/3
P
B 00
O
00 1--- CD
CD CD
O) CD CD
CD CO CD 00 ro
-7 in ro
in ro iri
ro ro rsi CM
CM ai in CM rsi CD CD CO CD 00 CD CD
1^
00
00
00 00 00 CD <D CD CD
r-- r-- ht~-
in
cr>
in CD
ai CO
o o o O
CO
D. CL 0. D. 0. D. D. 0. CL Q. CL D. D. D. D.
>-
> > >> > > > >>>> >> >
Z ZZZZZZZZZzZ Z z Z
CD
t
(D
o
oj
OJ
Q.
Q- >,
CO
-
.£ _1
E
O
c
O
<->
(D
H
n
O
O
—I
^
O
O
^
Q.
c/3
166
CD
>
z
Q. CL D. D.
r-^
>>>>
ZzZZ
r~-
r--
CM CM CM CM
CD
CD
CD
CIO
CD
ro
ro
ro
LL LL LL LL LL
LJ.
ro
CM CM CM CM
CM CM CM CM CM CM CM CM CM CM CM CM
CM CM
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
ro
LL LL LL LL LL LL LL LL LL LL u. LL LL LL
OOooooooooOoooooooooooOoo
oO
<M CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM
1-
H
_l _) 1- 1— _l _l 1- 1- _J _l (-
r;-9?
9?
co oo CO
CD CD CD CD CD CD CD CD CD
00 00 00 CO CO CO
in in in in
r\i CM CM CM CM
CM CM
CM CM CM CM CM CM Csl
C3)
12
CD
> >>
ZZZ
CO ro CM
CO oo oo 00 CD CD
ro ro ro ro
_1 _J
~
D. D. D. CM 0.
LL LL
E
-
00
ro in
00 (D
CD o in CD
CD
o o 00 cn CM
CM
CM
CD CD CD ro
CD o CD
ro ro ro ro
ro CM ro CM ro ro ro ro ro
CM CM CM CM
OO
oOOOOOooooOOOoOooOooooOooO
o
o
in
LL LL LL LL
CL
•
r--
LL
CM CM CM CD
00 CO 00 CO
55 O) oS o5 CO
r-r-- CD
ro ro co co
ro ro ro ro in
CD CD
9?
ro
CD CD
H
_l _l
CD
9?
CO
CD
CM
9?
5)
00
in in in
CM
9?
55
00
H H
CM
9?
55
oo
CD CD CD
CM
9?
55
00
CD
_l _l
H
1-
CD CD
00
ro ?5
CO CO
CD
00
?5
00
CD
go
CO
00
CD CD CD CD
Oo
o
oOOOO
CM CM CM
CM CM CM CM
r\j
CM CM CM CM
ooooooooooooooininininooooLOLninin
iv,fs,r^(-^rs_(v.t^t^f.^t^i^t^r^r~.r--h-r-h~cooooocooococooo
g;
>.
NIST NCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
<- CM CM CM CM
_l _l _J <- CSJ T- CM
Li. U. Li. _J _J
HH
LL LL LL LL 4"
=^
Q.
S
E
CQ CD CO
h-
m2
H
H
CN CM
Li.
LL
Li.
li.
CC Ct
OOOOOOOO^'^c^OOOO
OOOOOOOOOOO^^'^'^
OOOOOOOOmcDCDOOOO
to
CO
o
CO
C\J
CD
O
<- CM CM
CD
CM CM CM CM CM
^
O
CO CM CO CO CO h- hCM CNJ (M
OOOO
<u
=-
0
"=1;
CD
ro
CD CD CD CD
O
m
CD
CD CD
in
0
n
in
in CD ro CO
q:
C C C c c c c c c c c c C c c
CM CM
CN
CNJ
CM CM
^roc30oococoi~-rocNcocMCDr~-Lnr~CMrM'<--<-'<-T-CMCMCMrMCM'<-T-T-T-
u^cD-^(~-cj>-^OOocDt^r~-roc»oro
Ln-^CMCMCMCOCDLnrOCMCOT-T-T-T-
CO
0OCOCX50OCX5CCI00COCOCOCO0O0OC»0O
CO
or-LncDCMCDi^-^CMCJi-^cjicDincn
^cNicTicrjcDtri-^'^TCMCM'^'-CMCMCMCMCM
—
o > ^
°
cor^-CMincjicororocDco'rfinin^-
"J" t/1
cnTrtoinmincocDincD'^-'^rroro
on'-Ln-^ocDCDr--cDa)OcDC3)
cDoocricricDcrjcdcrjcncDr^odcDcb
o
T-000'^--'-'-0'^0000
^
Q.D.CLCLCLCLQ.Q.D.D.D.Q.D.CLD.
>>>>>>>>>>>>>>>
.£
Q.
CD'— ON-OCDCJ)cr>CDOCOrO'*COCJ)
"3
rs,
oocnocr>ocnoorQ'*cocMCMCM'>-
cDCD'^rrO'S-nt^t^cDcncjirororon
< - OO'-'-^'-OO'-'-'-'-'-^--'^
eOCD<3C3CDCDCDCDOCDCDCDCDC3C>
CD
CM CM CM CM
o
0
Q,
Q.
Q. >,
E
o
c
O
15 15 15
LL LL LL LL
00
Fla
CM CM CM CM
"o "o 15 15
CD
O
o
15
Fla
I
Q.
E
000000000000000
o CMCMCMCMCMCMCMCMCMCMCMCMCMCMCM
o
QJ
O
m
CL
CD
-I -1 -I Ii2
</)
o
^ ~C 9P ~C O 5c >
CD
cu
•2
15
P
o
iii
—I
o)
I o
i o
o
o
H
-J -I
oococooooococooocororor~-r--i-^i---
oocogooocooooocO'^-<s--^coooooco
LnLDLnLncDLnLnLnoooLniDLnLn
COCO0O0OCX)0O0O0OTf-^T3-0O0OCOC»
'-^^^T-.^-.-T-cDCDCDOOOO
cnLnLnLncoLomLnooooooo
^•^'^^•'^'S-'^-'i-'a-'^-^rorocooo
T^T^T^T^T^T^T^T^CMCMCMCMCMCrslCM
—
£
CD
•a
^ LL
1Q. 2
c/)
OOOOOOOOOOCD 0000
C35CJ)C350)CJ>CJ>C3)C35C35CX
OJ
1-3D,
WTC
Investigation
c
o
q3
c«
03
^CD
OS
Q -r
.
NISTNCSTAR
o
^
JJ
167
Appendix A
(0
x
(Q
<N
CM ro
(M CO (N
^^^^^COCOC/)COC/)COC/)C/5CA)C/3H_l
c
o
C0CA)C0C/pC0^^^^^^^^^^iO^
I— KHt— I— CQCQCQCacDrororororoD.D.
Cl
E
TO
CnCDCDOICDtacOOOOOOO'.—
.—
T-T-T-'--'-
C0C0C0C0C/)ZZZZZCA)CA)C/)C/)C0OO
J3
CD
•D
000f
OrOCMCMLDtDOCMCNJ-^OCSjrO
CO
CNJ
CO CD CD
m
CD
O
r--
T-rO'^CNJCMCNCMrMCMCMCMCNirNJ
CM
c c c c
CNJ
c
c C c C C c
CM CM CM CM
UJ
CO CO
CD
CD CD CD CD
0
c:
CO
00 CD CD
in
c C c c
CM CM CM CM CM
r\j
0
CO CM
CO
CO
00 CM CM CM CM CO CM CO
CO CO CO CO CO CO CM
CO CO CO CM CM
X3
CD CD CD CM CD CD CM
CD
CD
CD
a> CD
T— CM T—
CO in CO
CO
CM CM
1^
in in in in in in in in in in
0
^
^
CO
CD
CO
^' CO CO Lri CO
CD CD CD CD CD
lri
lri
r--
CJO
06
06 CD
h- hr-
lri
iri
0 0
O
c
ro
o5 :g -5,
CD
0
in co
CM
CD
CD o> ID CO in CM CM
in CD CD 00 CO c» CD CO CO CO
CD CD en
00
CM CM rsj CM CO
CO CO CO CO CO CO CO CO CD CD
CO
CM CD CO
CD CD en CD 00 CO
lri
CM CO lri lri CD
CO
CJ> d> lri
in in
ro CO CO CO CO LD in in in in in in
CCD
O
Q. Q. Q. D. CD Q. D. D. CL D.
>>>
ZzZZ
CO
00 in C»
CO CM CM CM
cn cn cr> O) CT)
CD
CM
CD
m
i;
CNJ
> >> >>
ZZZZZ
CD
03
E
o
q0
Q. Q.
>>
ZZ
in in in in
00 CO
CO CD CD
0
00
00
in
00
00 in CD
0
0
0 0 CD
0 0 0 CO
CM
CM CM CM CM
0
0 CD
0 cn
00
CD CD d CD CD CD CD CD CD d> d> d)
C35
C3>
cd"
"D
Q>
in
>
03
CT)
C35
CNI
CD CD CD CD CD
CD
ro
c
D)
g
CM CM CM CM CM OJ CM
11
O
CD
DCO
CM
r\j
CD
CD
CD
U- LL LL U. U_ u. U. U- u. LL
LJ.
CD
-D
CD
CD
CD
CD
CD
CD
CD
CD
CM CM CM q: a:
if) CO
CD
CD
CD
"cD
ro
CL
O
0
0000
000000000000
CM
CM CM CM CM CM CM CM CM CM CM CM CM CM
CNJ
CNJ
CNJ
0)
I- 1- -J -J
E
ro
c
c
CD
o
9
"o
S
C
o
LL LL
CD
i
XX
L. LL LL LL
c
CO
Q)
T3
i 5 ^
8
" c
E
i
0000000000
03
m
CD CD on CO
in in
CD CD CD CD CD CO CO
06 06 06 06 00 CD CD
CD CD
COCOCOCOCOCMCMCMCMCM~~~~
CM CM CM CM CM
CM CM CM CM CNJ
E
o
03
Cl
0
0
CO CO
CNJ
CM
DC
ro
Cl.
CD CD CD CD CD in in
168
NISTNCSTAR
1-3D.
WTC
Investigation
Data Tables and Supplemental Figures
03
0)
03
0
c C c C c
03
03
Ol Dl o> D) D)
_ra
_ro
_ro
_ra
_ro
O
0)
D) D) CD
flange
web
web
X X X X X X c c c
o o o o o o _ra TO
0)
03
oO
'a-
CO CD CD CD CD CD CO CO
LL LL u. u. L. LL LL LL
§§§§§§§§
r\i
CM CM CM CM CM CM
-a-
CM
_l _J
CD
CM CO
L3
LU
in _J _J _l _l
CM in in in in
§
m
LLI
0)
03
03
1
1
§2
CM CM CM
CM CM _1 CM CM CM
in in in
in in in in in in in
CD CD ID CD CD CO CO 00 CD CD CD
CO CD
XXXXX
03
CL Q. CL CL
>. >. >s
1- h- 1- 1r— r—
CM CSJ CM CM
_j _j _j _j
LL LL LL LL
o
OOOOOOOOm
0)
CO CO CO
CO
CM CM CM CM CM CM
LL LL LL LL LL LL
03
§§
CL CL
>> >.
1- 1in CD
LL LL
CM
CO cn
O in
o CD
O
CM
ii
LL LL LL
CM CM CM
in un in
CO CO CO
CD CD CD
03
5
CD CD CD CD CD CD
CM
O
O o C3
CO CO CO CO
OOOO
CD ci
CO CO
(J
O
o-^co-'-'q-'-'^incocO'^cDi^t^'g-CMTCM'a-ococM
COC0030CT>COC7)COCOCJ>0000'-OOOCMt-OCO
m
00 in
CD CD LD in "S
CM
CD
CO
CD (O CD CD CD
C7>
CM CM CM CM
DD
COC^CTlin-^CMOO-^TtCO-^
CO
C73
in in in
03
03
to in
o
CO (D
"E"
o
in (D CM CM
CO CO CO CO
o
"c
o
"E"
cn 00 00 OO CO
CO CO CO CO CO
o
CO CM CO CO
C33
CO CD CM CM CM CO CO in
q"a-
00
00
CM 03
CD CO
in in in iri CD CD CM
CM CM CD
CD CO CO CO CD CD CD CO CO CO CO CO
d
o
03 CO CD CD CD
CM
CO in
in CD CD CD CD in CO CO CJ> CD in
(N CM CM CM CM CM CM CM CM CM CM CO
o
^
CO
CO CM CO CD CO CO 'a; <J>
CO
CO CM
CO CO 00 CO CO
CO cr> CO CO CO CO CO CO
CO co
CO
"5 .5 9;r:
o
>-
^«
a.CLQ.Q.Q.CLD.DLaCL
OCT>CJ)CT)00C0C0f^
D.
>
CO
d
CO
CO
d
CD 53;
CO
(d
CO CO CO
CD 1^ in
O) CO
CO CO
CO CO CO CO CO CO CO CO CO CO CO
CD CD CD CD CO CD CD CD CD CD CD CD
03 CJ) CJ) O)
o>
O) 03 03 (33
oooooooooo d d d d d d d d d d d d
-o CD
T3T)-0-0-0X)'0"0"D
(11
CM CM CM CM CM CM CM CM CM CM CM CM
XI XI
T3 TD XI -D
CC o:
cr d: a: d: ca CC q:
Da
XXXX
a
H
O
cna)a)Cj)Cj3Cj3tj3Cj3a3incocococo^z:
COC000COCO00 00 0OCOCJ)COCOCOCO--,-II
SSS<oSSSS5cMcoco?5co|£?if2
CO0000COC000C0 00 00CJ3COCOrOCO
,
,
00 00 00 CO 00 CO
ooooooooooooooSS
oo
o
ooooo
0)cno3a)03 03CT>0)0)coinininin'^"-'
tr-4-^^-^-g--4-^^coTT-g--4'4Gp3 o o o o
CM CM CM CM CM CM
NISTNCSTAR
1-3D.
WTC
Investigation
169
Appendix A
c c c c c
(U
0
CD
CD
0
(T3
o o o
en
CT3
CTS
CD
EEEEEX2X1XI
CO m m
OOOO'^'-'-'-"--^
(35CJ)CJ)CT)C£!CDCDCDCDCD?S2S'^'~^f^<~^'~^'*'^'^
— — — — — — — — — — U-iL-,U-,cDCTic7)CT)c:nc7)(j)C7)
§§§§§§§§§§
CMCMrMC\]CMC\]rMCNICMC\i
^U-LLU.U-LLU.Ll.Ll.
S^S^g^g
>%>s
§§§§§§§
TTT'TinincnuSinLnOOOLi.u.Li.u.u.^^^
<— ^t-mcnLnLr)mLnoocDcoi''''cDCDCD
r^t-~r~-'-^-^'-'-'-oocDcoIIXIICNic\]CM
OOOOOOOOOOOOOIIXIIOOO
cDoO"*'-cDcDrsi'-M-cDooLnr^coocJ)'-'a-CTiOoo
"d-cj5a5cn'a-M-CM^T-^(\jr\icNro^c\jCN(J>ocT)
r^,-^(DO-^''^"-r^'^-^^'^'>-'^-^'-'-'>-'^ci'^C3
cncocj>orM(NCT)'^cj)nc3icD5ooi^cx30ocDCDco
cniniricocDcninLn"*Lninin c'DCDcncDi^cDCDco
cccccccccc
cNicNCNCNCNCNir;iCN(>jCSCSpJ^£SES
cDcocoi^ocDccO'^cJi'^,
cNCNiroronrMcsjromcNj^^
'-cMcDcDt^cr5cncDi^r^co-^oO'--^0(Nj-7-iOco
t^r^corocDcncD'^"d--<3-CT>ooir)cnrs]CM'— ->>—
^'^'^^^^•^LnLom'^LnLOLOuocoioun clouo
n'^r\icDt^c»'5trMTt'^r\ir^cr)cccDcD-^rM-QO(35
ci)cbc^roO'-ocr)CjicT)C\icNCNic3icT)uriin"a->'a-^
0)
cDcDCDCDi^i-~-t^r^r^t^t^r^i^r^r^i^t~~r^
a)
xf
:g "5,
03
o
^
2^
a
o
-5
"c
p
IT)
CO
rsi
ID in CO CO
TO
.9
O
o
CD CD
CJ) CO CD CD
CO CM 1^ CT> CO CM 00 CM 'J CO CO CD
in in
CO CD un 00 00 CO
CO
CD CD CM
CO CO
Cvl CM CO CO CO CO CO CO CO CO CO CO CO
CO CO C\] CM CO CO CO
CX3
cr>
">
'-t^r-^
00 CO
ro CM
CO CO
(J)
ro
CD
CM
in cn
ro
c
o
^ o
^- cn CO
csj CO CO
CD CM
in CO CD en CD en in
CO CM
CD
"3-
CD
CO
CSI
CJ)
YP YP YP
Z
Z
CM
d
in
CO CO CM
YP YP YP YP
CD CD CM CM
LO en CD in
z ZzZ
<-T-^-.--.-T-T-T--^T-T-T--.--^-^ 0(MC\JC\J
OOOOOOOOOOOOOOOOOOOOO
E
TO
nj
o
(J)
g
w
>•
£
2
E
o. Q-
w n
CM CM CM
r\j CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM
"D T3 T3 T3 "D T5 XI XI XI
ro ro ro
cr cr rr cr cr cr cr cr tr cr cr cr cr cr cr cr cr
LL LL
Flat LL
CM
DD
XXXXXO
ooooOoOOOooooooooooo
O O
CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM
_l _l _] _1 _] _l _]
ooo
ooooo
c c c
o o o
E E E
ro
0)
ro
CD
ro
CD
XI X) XI
m
P
E
ooooooooooooo o o o o o O O O
in en un un in
CJlC3:iCJ)CJ)CJ>CJ)CJ>CJ)CJ)CJ)0O0OCO CD CD CD CD CD 1-
1-
§
0
170
—
o
oo
in in un
NISTNCSTAR
1-3D.
WTC
Investigation
Data Tables and Supplemental Figures
CO
z
CM Z
CD
CD
CO
Z w<< <
<
1— 1-
•7
CM
CD CM
Q.
E
^ CM
< <
CO
<- CM CO
un CD
q:
cr a: q: cc
I- H- I- I- I-
C/5
ro
in
ro
< < <
t—
CM
CM CM
CO CO CO CO
<
< <
C» 00 00
O
H
CM
CO CO CO
ro CO CO
LD in LD
OOUo
0
T3
T- in
in Ln
T- CO
CM •<- CM
O
'5;
o
UJ
O
<-
CD
o o CO
o
O
O
CXD
ID
CM CM CD
CD
CD
Ln
CM
O
LD LD 1^ CM
CO
CD CD CD CD CD CD
o
CD
CD
T- CD
Ln CD CD in CD
^
CD
CMCMCM^^I-
>-
.-T-'.-.^T-
t^->d-LDCOLDCD
CMrM'<-CMrocM
CO
ro
COt-LDCMO
T-r-TCMCOCM
rororo
CD CN
00 CM ro
CO
CM CT> CM
in in en ID ID lo
CD
ro
CD ID <- ID CD
TT- TID ID ID ID in
CD 00
CD ro Tro
CNJ
cororororo
O
55
CM
CD CD
^
00 CD
LD LD
•<-
c
^'"CO
CD-^^
OO^rO^
ro
ro
rsjCM
00 "o CD
CM
CM
ID
ID in
^
ID
ID ID LD
ro
LD ID
C
CO
00
r~-
O
ID
1^ rCO CD CD ID CO
00
CO 1^ 00
^
ro
CD
'a-
ro
ID
LD (3^
CD CD CD
h- r--
t-~-
O
ro ro
LD
00
CD ID
CD CD
CD
m CD
1^
E
ro
1
CO CM
CD rsi
ID CD ro CD CO
ro
ro 'd>a-
LDh-cD.S3 LDh-rocD
o
o ^
ro
ID
CO
CM
O
.2
oooor^
roiDoo
in
ro CO ro ro ro
oooor--r-,CDrocDCD
rororo":ro-<d-roro
ID
ID CD
CO ID csi CO
CM
ID CO ID CD ID CO
ID CM
00
CO 00
cr> CO
ID
ID in in
CD
CD
LD CD
LD LD LD
ro
ro !^
in CD Ln in
ID
ro
CO
CD ID ID
Q. Q.
Q. D. Q. Q.
Q. Q. Q.
CL 0. Q. CM
0. CL CL
zzz
(D
in
^roro
0
q
CO
03
^
•c
o
>
Q.
<
Q.
>> >> > >
ZZ ZZZZ
.E Q-
-
>> > > >
ZZZ Z Z
LD
00 CO CO LD
CM CM CM 00 CD CO
CD CD CD
CD
00
OOO
O
d> d> d> CD d) CO
-2°
o
oOooO
> > >
Z Z Z
> > >
ID CD
CD CO
ro 'a- «d- ro CM
CD CD CD CD CD
7^
CD CD CD CD CD
re
Q)
> > >
zz z
0
0
CD CD
0
ro CM CN
0 q 0
CD 0 CD
CO
ro
00 00 00 CO
CM CM CM CM
ro T- >ro ro
-"dCD CD CD
qqqq
CD 0 CD CD
q q
o o o
g'q
CO
c
c
CM CM CM
.E _,
-
S
Q.
9>>
E
o
C
0)
DD
O
D
T3
Ql
q: CH d: cr
1_
o
ai
to
Q.
o
OOooO
CM CM CM CM CM CM
CD
CO
£
Q-
s
0)
2 D
TO
O
o
—I
CD
o
-
Oi
C
O
m
CM
to
(0
Ll.
CO
ID
LL 01
oO
OO O
CM CM CM
O
CM
r\i
£^
_l _l _l _l
TO
:
CD
u. U. U.
ro
ro
"c "c "c
ro
ro
ro
^^^
(5
D
^^
ro
ro
ro
~c "c "c
(A
O
tn
—
(/)
O
o
ooooo
LD ID LD LD LD ID
3
1-3D.
U)
WTC
Ifl
(/)
3
Investigation
c
<
Q£
CD
ro
(/)
(/)
3
CM CM CM CM
0000
ro
Z
S
2
»^
c
00
ro
ro
ro
ro
LL LL LL
3
0 0
0
CM CM CM
E
_i _l _J
Oo
OO
o
^^^
r--
r--
CD CD CD
CM CM CM
'd-
o
a
X
c
E
0)
I-
E
o
o
CM
-
ro _ro
_CD
^
_ro
c
T— T— T—
ro
O)
ID
O
SS.
O)
c
c
O O
O
LD Ln in
_J _l _l
CO
ra
k_
0)
O)
3
ro
0)
k.
(A
(A
u)
ro
0
000
CM CM CM CM
</)
<
ro
LL LL LL LL
CM CM CM
_o
O)
Si
c
o
ooo o
LD LD LD LD LD
u
o
o
oo
CM CM CM
fl)
D
CD
NISTNCSTAR
E
o
ro
V)
(A
0
ro
.i
3
00
^
o
o
TJ
u. U-
c
ro
ro
.
CM
o
^ >
9 ^ P ^t —O
CO
^
X
c
_ro
Ll.
in
o
c
o
_ro
_ro
_i _l _l _l _i
is
2 ~
m
C
o
o
(A
OOO
ooooO
.It:
<
(A
(/)
3
OOO
O
LD ID LD LD
<
to
(A
CD CD CD
CO ro ro
3
17]
Appendix A
CDCUCUCDC1)0<DCUCPCD<D
CD
Q)
CO
EEEEEEEEEEE
.
V
rsi
V
CO
CO
^
t-
^r-'-H
^^,-^^T-^(JQQCO
CX)COCX5COCDCOCT)O^C71<^2^
Li.
>,
CD
O
O O O
CD
< <
CO CO |Z
^ 5
S S ^
5 2 O
h-
CI)
o
CD CO
CM CO CO CO
r\]
CO CO
CO
o
oo
^
CD CN in
1^ CM in
Q. CD
in CD CD CD CO
M-
^
CB
CD
CO
cji
CQCDCDCQCDCOl—
CD
0)
cn
in -H
c/)COC/)C0COC0c/)HI-|-|-J-'^
£
CD
CD
in CM
in CO
o
O
csi
r-~
O
CT>
in
CO
in CD
CO
o
O
CO
CD
D.
o > ^
oO
CO
CD CNJ CO
CO 00
CO
iri
ai CO CD
CD CD CD CD CO CO CO CO CD CD CO
ooo
1^ C\J a>
CO
CO in CO
in CO CD CD
CD CO CD
CO CN CSl CvJ CO CM CSi CM CM CO CO
CM CO CD
CM csi CD
CO CD CD
o
CO CO
CJ) CO CO
CO
CM
CO
o>
2
^
CO
cri
in
CO
00 CO CO
CO CO CO
CL in
in
_^
c
>
Z
p
CD
00
M- CO
o
ooo
^^
CO c^
NCJ)
CO
CO
d
C7>
o
in CD
c\i in CO
a-
Q. D. cn
CD
> >
ZZ
in
CO
CO
in r- CO
d
csi
cn in O) O)
O)
CD in CO
00 CO 00 CO
CJ) 00 1^ CO CD CO
3"J
'd- "3•>a'a-
ooooooooooo
o o o CD d CD CD o di CD CD
o
CD CL
00
in
CO
>
Z
(3) CO
00 00 00
CT)
"3-
o O o
d d d
T3 (5
CD
"to
•D •o •D
tr
IT
^-
D
a
T3
o: q: cr
-o •D T3
cr
q:
oo
ooooo
oo o
Oo
CM CM CM
CN
CM
CSi
0)
"Co
g
D
D
cr cr cr
o
CNJ
o
O o
CM
CM
Csi
^
OOOOOOi!,r\JCN
oooooo^oo
CT><DC3>C7lC35a5^Cr)C3>CJ) c
rsirsicsiCNjCNjcN^iobooco-^
m
"D
E o
Q.
CO
172
0
>-
NIST NCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
(Da)(D(l}00(D0
0 o o o o o o oi5iPi5i5i5i5i5i5
'^'^-H-M—
mmmrnmmmm'*-**-'*-**—
a>0(D<B<D<l)00>>>>>>>>
Q. Q. CL Q. Q. Q. D. D.5 >>>>>>>
>.>,>.>.>.>.>,>,C\IC\]CN(NC\ICMCNC\J
l-h-HHHHHH-<-^T-^,-.^,-TT-
csj
n
LD CD
il^COOlO^CNO-^
.TV
LLLLLLU-LLLLLLU-VVYi
.--^•.-r-'^CSJCNI
1
I
•
1
1
E
1
I
I
I
1
I
CNCNCNCNCNCNCNCMv!--A-A-At-I--A.Av^
TO
LDLDLOLnLnmLomi
i
i
i
i
i
mcDCDGQCDDDCDCDOOOOOOOO
oooo-^-^"g"g-^v-pppo"g"£
d CD d d d d
dddddd
"D T) "D TJ X)
;a :o :u :u :u :u
"£ "S "c ~c "S "S
55
:5
:5 :5
"£
"H "c "S "2
p
cj)
CO
CiCi
oaD-^t-^inu^-'-O'-ocof-inin'-CN
CNJ(J)T3-ncOCO(DlDCNCSI^OCT!COI^t--
•55
6^
Sj:nc\i^TfcDcbdcTigg'^2
^
^ifCTJCT)COC0CO(DCT)CO^'-^
^
^
at
/«
r~
—
)
——
CDLO
—
U3 CO
-5
"£ -5 "£ -5 -5 00 CO "£
"c
CO CO
5
5
;c
o
TO
Q>•
c
5
"£
-5
"c
O) O)
CT>
m
c
a>
cr>
— — —
^
TO
TO
C
~—
-CDtD
CO
CO
'c CO
CD ID
52
CO
dd
CO ^
CO
CO CO
TO
-3
"c
C35
CO 00
(O (O
c
"c "c
CJ>
CJ>
CJ)
CO CO
CD CD
CJ>
00q000
00000
d0dddddddddddddd
T3
O
n0
mmm mm
q: q: tr ce ir cr
z z z z z z
ffl
Q.
o
E
'
0)
to
$
2
(D
o
CM CM
-a X)
tr cc
m m m m CD CD
tr cr
ce
z z z z z z
CM
IT
0
000000000000000
CM CM tN CM CM CM CM CM CM CM CM CN CM CM CM CM
_l _l _i
_l _l _J _J _l _l _l _1 _] _l _) _l
D.
illc p o
C
§ >
•~
CM CM CO CO
rs]
c
o c ._
°Q. >. E
o
c
COCOCOCOOOCOCOCOqqqqqqqq
O)
O
E
o5
:g
CO
o
i
c o
^ZC^Z^Z^ZIZCCcocococoooooooco
a;
^
5
I)
o >,
0)
COCOCOCO0OCO0O0O£j,(j,(3,(j5(j|Q,|-„Q,
_
>•
NISTNCSTAR
1
-3D,
WTC
Investigation
173
Appendix A
a:
a:
T- CM 00
N
<
X
a;
CL
E
1
CD
CD
N Ni N
< < <
X1,1
XX
CO
c
g
o
U. LL u. LL
0^ 01 CC 01
CO
I
t
I
(U
m
m m cB
CO CO cn CO
O (J O O
CO
c/)
w
00 CO 00 00
UJ
c
0)
c
o
0
in CM
:£
a
00
CD <CO
ro CD
CD CD CD CD
CD
00
CM CO CD CD
c\i LT) CM rsi
(3) CT) CJ) CD
CM
ID
E
o
u
DO
^
CO
i_
o
Ln
o
"4-
00
CO CD
>^
TO
(/)
CO
3
(/)
—
0)
L_
o
C/l
in
r-CD h- CD
•*-•
V)
0)
CO
o
2
Q. CL D. CL
O
>> > >
>-
CL
a
CO CO CO CO
0)
^
< -
'd-
d
CD CD
CL
>
'd<d-
d
o
o
E
o
TO
(D
QJ
I
<
Q.
Q. >,
CO
.E
E
O
c
Q
o
H
^
O
O-
E
O
NNNN
< < <<
XX X X
CD
CM
o o o
o
CM CM CM CM
O
CM
SI
nj
O
CM
^ ^ o
m
-
-2^
c
0)
c 0 F
^ >
u
O
OO
OO
OO
O
If)
I*
~-
— —
(U
Q-h^ h(/)
g
TO
u
—I
12
CD
o
E o
0)
C
CD
CD CJ)
CM CM CM og
C7)
-sT
_o
u
CD
CO
174
o
«
a>
—
><
^
<^
^
0)
c
a
0
E
"o
Q.
«3
^
o
c
(O
CC
EC
^^
CD
"c
CD
^
(0
0)
X
NIST NCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
CM CO
Q.
CD
E
z
CM CO >-
KM
i
3
i=
•<a-
J5
"c
c
CA
> >
IT) to in
in ID Ln in
CM CM CM CM CM CM
<i-
CN
-!5
CD CO
c
0)
E
o
CM
a
CM
c:
CNI
CM CM
02
CM
(0
O
CO
Ci Ci
CD
CM CM CO
"i5
*•>
0)
O
E
O
CO in
CO
to in
CM
o
h~
in
in
o
I
(0
O CO
CO CO
OO
CO
in
CO 00
in in
CO
CO
in
CD CD
CD
o
00
o
iri
CD
o
^
-n
cn en
3
(A
(U
k.
ro
O
CM CD
CO CM
ro
o
C/3
*^
OO
(A
d >
CO CD
oS
in
^
iri
>-
> >
ZZ
05 Q. CL
iri
CD
00
CD
>>
Z Z
0628
0995
0984
CO OO
2^
ro CD
CD
iri
(0
c
Q.
o
Q.
>>
Z Z
CL
0)
3
(0
k.
0)
Q.
0995
0660
0995
,0621
,0995
oo
oooooo
E
-o
c:
(0
E
o
o
O
CM CM CM CM CM CM
(0
O o c
<B Q.
Q- >. E
C/)
-
Flat
Flat
Rd-
o
E
03
o
o
Rd-
Rd-
Rd-
01 a:
oooo
o
o
in
o CD
CM o in
CD CM
o
o
CM CM
I
<
a;
^^
™
~£
^^
E
QD
T3
0)
coin— c^^f^TO*^*^
o
.2
o
o
—I
o)
inin
I
Q.cc'ccc'c
Q.
3
—
CD ^
—o
I/)
o
'o
k_
0)
1 2
CO
o
a
>>
NISTNCSTAR
1-3D,
WTC
Investigation
175
Appendix A
A.2
DATA TABLES FOR HIGH STRAIN RATE PROPERTIES
The following
176
table
summarizes the high-strain-rate properties.
NISTNCSTAR
1-3D,
WTC Investigation
— — —
Data Tables and Supplemental Figures
in in
CD CD
^IV.
in h-
N.
r-~
t-~
r--
1-^
t-~
s.
r
CM CO
ooooooooooooooooooooooo ooooooooooo
CN
CSi
O O O
d CD C3
m
o o
in
CD CD
CD CD CD
CD
CD
CD
_l _l _i _1
O O
CM
_l _i
_
LL
a: cr _j cc
LL.
a>
o
Li.
C\J
C\J
H
CM
h-
m
12
m
_i
hp
LL uL LL LL LL LL LL Lll LL
(T cr cr 01 IX ce cr IT cr
c
E
3
O
o
L.
o
O O OO O O O O O
0)
'O "O
O)
<B
E
c
TO
ro
LL LL
TO
-D3-
-D3-
6 6
OD CD CD
O O
CO 03
z z
^
O O O O O
is:
cn
-RF-
CD CD
t- I- t- I-
.£
c t
P c
a,
^
C C dj
o
1
1
RF-
LL LL LL
a: 01 01
OlQlQlOlOlOlQlQCQl^
CD CD CD CD CD CD CD CD CD
0)
—
1
RF-
01 01
cr cr 01
m
o o o
CD CD
tf)
o o o o
d CD CD CD
CD CD
CJD
I- I-
i5COC/3COCOC/DCOCOCOC/)CO
j^JpcTJCCCO
C^CXJCTJCDCUCDCTJCDroCO
C/3
CJl
(J)
cj>
cyt
is:
is:
i^:
Cl
01
CO CO
Z z Z
H H
LlXXXXXXXIXX
c/5
Z Z Z 2 Z Z
H
M
X X X I I
CO CO
cn d)
CO CO CO CO CO
if:
H
CO CO
O O O
LL
X X X
c/5
X X X
X X X X
CD
(T3
CO
CU
TO
CO
TO
TO
LL LL LL
a
m
comcO'^-r-^r^-^CNj-^roror^co^cDCDOCMCJ)0-^ooc\ir^
cDcoininmmm-q-rOM-'^'^'^cDminmmiOfrjron-^mcocO'S-'a- m o
m o o
T— T—
-a-
(0
CN
"5
(5
o
"3
0!
CSI
>->-
-r-
(0
O
(0
0)
^
o
cDmco-^cDcsiooooc\jr^cocMincDomcO'*'a-cDcO'5i-c350
CM
CO CO CO
Uj
o-^co(jicMCDcDm-<3-ncor\jo-<-concMcnt--cococ3icocDCMcocooooo-<—
(0
c
©
r~
^
r~-
m ^ CO 05 ^ CO
CM
mcD^cjjggg^^cDS
COCOa5O>^^~00COC7)^
X
n ^ CD CN n ^ O ^ O
^g^gc3.c6c3ir-gg CO CO S ^
!I
oocoTt
^
O
^-f
r--
0)
OOO
c o
C/5
i-^co
1^
CM
O
o
m
r-
CO
CD
h-
CT)
rCD
CO
CO
!E
in
'
C35
.
rT^
coco
OO
nco
OO
rri
c-ri
5 o
'
CO CO
CD CD CD
t_>c_)
o
UJ UJ LU
CD
1^
CM CM
O
TO
tr
3
re
cDmr^cocococDCM-^ coi^r---*r^mmcM-q-r^
0)
a
COCMCOCDCMCMCOCOCO ocor^cocMO-^cocD-^
^<3--»-<3-mmtn'^-<T inininiD-^inininmm
^ —
-.-
CO
h~ CD
CT>
•<t
in
m
CM
m
CO
CO
•<3-
O CM
o
in CD
E
O)
cMr~-cocMCTiC350cDm-<3-c35r^oocMomcoocoi-co^
E
o
o
^ —
C/3
^dcMmmLricMCMco^-cocMCMcocIi-r-r^lncc)r--^^n'a•
SS
2 °
o
LU LU
>- ,CD CD
CD
I
p
(T,
Ln CD
O
CD
CM CO CO
m
CO CO
(D
a
2 O
^
^
</>
m
o
m
o o
o m
m
CO r-- UJ
CD
CO co
CM CO CD CD
CM
o o
o c o
=
i/>
p
CO 1^ CO CD
CO
CO
-aCO cn
CO
in
CD
CD LU LU
00 00
CM CM 00 CO CO CO
r-~-
55
ca
JQ
m
m
CD CD
in
o
o
O
CD
T- CM
oi
X
d
^
_aj
<
CM CO CJ) CO
CD
CO
1^
CO
CD CD
l-HH_i_i_J-i-i
_l _i _l _l 1- 1-
_l _i 1- 1-
<D
H
t-
:^
CM CM
o
o o
O CM
O
o CM
o CM
O CM
o o
o o
o o
o o
o o
o o
o O
O O
O CM
O CM
O CM
O CM
O CM
O O O CM
—
P
.2 CQ
"S
o
CD
•
S
S (TS /T^ fTS /tS /T%
(^^ ^^^
??9???9?9??9?CTicj)cj)cricjic35cr)cjicr>c35C35Ciici)CXicT>cj)CT)05CT>cT)CJ)cn
CD CD O) CD CD O)
cn
CJ5 cp CJ)
~ ~ ~ ~ ~ ~ ~ O O O CD CD CD
CO CO CO CO 00 CO CO
CM CM CM CM„ CM,
,
, ^
„
qq"31^
r-^
o
1-
,
<U
^
<D
>
c^
o5 55 35 Si
35
•
'^^
>
>
V
>
rt
CD
2
<u
—
NISTNCSTAR
ooooooooooooooo o o o o o o
CD CD CD CD CO CO
1-3D,
WTC
Investigation
mm m
m
CD CD CD CD
m m
in
CD CD CD
111
Appendix A
ooooooooooooooooooooo
in in
in in CN CM in LD cNj
CN r\J CD ID CM CNJ CD
T-
CD
CM
CNJ
IT)
en
csi
CN
CSI
CD
XI
_i _i L2
L6-
o o o
<3
-L4
RF-
RF-
RF-
RF-
RF-
L2
RF-
L7-
RF-
L3
RF-
03
a.
E
LL U- LL LL
CM
h-
r\]
LL LL LL
a: a: tr
LlL
h-
LLLLU-t-QCDoaoQcODDCOCDCDCDinCD
^^mcD22cDCQCQ^^^^cocr)rorOfOO->roro
CMCMCMCNJCMCMCMCMCNiOOOOOOOOOOOO
cmcmcmcnjcmcnjcmcnjcmcdooooooooooo
cncmcnicncnjcmcmcmcm'^-^-^-^'^-^-^'^'^-^-^-^
^
—
O t
C
fc
O "D
(D
c
Q.
^ ^
H
z.
-z.
I- I- I- I.£
O)
CNJCNJ
CSICNJ
(u
roj5c/3c/)i5i5c/)(/)coW(/Di5roc/)c/Da3c/)iPiPc/)(/)
LlXXLlLlXIXXXLlLlXXIXLlLlXI
CD
TO
CD
to
CD
CD
CD
CD
CD
OOO'd-OCNCMLO-^a-CSl
(D
^
"3
TO
O
g" Uj
>a-t^cj>oot--T}-ocD
(D
CD
CD
CD
(D
(D
^LOc\]Lr)oocx)U*)ocDr^cJ)
WcncD-^cD'S-cD-^coron
CMCM-^-^CNJCM-^-^
~
CsJCM
CsJCNJ
0)
rvJCM-r--<--^-^CMCM-^-^
rMCMCMrMT--^cMncs]co
CO
o
CM ^ ^
(X)
O
03
cDCT)CDc\ji^ro'^ococn-^0'-cMco->a-m-!-t^Lnrri-'^cocT)-a-"^rocncO'3-cnroroOLn'a-cDnc\ja)CM
CDCDr^t^<DCDI^r-COCOCOCX)00CT>CnC3>(Ji<X30O0O(35
^^oc35(^j^cr5cpcT)Cp'i;-rfincqc35C3>ocncj5cq'^
^^^-^-^JjJJj^O^CMCMrMCMCMrOCOn-q-CNJ^CMrO
^5
in in
in
o o
IT)
c o
K
tr
o o CO
o
o o
CO
C3
LU LU LU LU
CO CO
CO CO
CO CO
O o
CM CM
CO CO
o
S ° uT
-fc
SS
^
0)
CO MCO CO CO ro CO CO CD CD
(M CD CD CO
1^
in CM 1^
in CO CO CD CO CM
ro
CD 00
lO in CD CO in in CO CO CO r-r- CO 00 00 00 r~- N.
CO
co CO
1X3 CO
t--
.9?
S o
in in
CD
o
UJ
CO
CO
CO
cn
LU
CO
CO
CO
CO
C35
o
CJ5 CJ> OvI CD
CO CO CM CO
CO CO 00 CD 00
CJl
CJ)
in in
o o
CO
n CO
in LLl
CM
CN CO
CO
CO
o
in CO CM
CO <D CO CM CM iri
CM CM
CM
o o
LLJ
CO
CO
^
in in CD co
CO
CO
CM
CD CO
in
CM CO
CM
rsi
-q-
—
CD CD
o o
CO
CD
H H H
_
Q.
« g -
o
s;
m
^
S
o
o)
:g
.9
— m
I
o
1<i>
>
a;
-Q
0?
0)
E
<u
=j
r^r^t--!^r^l^^-l--t--CDCDCDCDCDCDCDCDCDCDCDlD
ininunininininuninoooooooooooo
->--^-<--^-i-->--<--i-->-CMCM<MCM(NCMCMCMCMCMCMCN
-A-A-AT^-A-A-A-A-^CMCMCMCMCMCMrvJCMCsJCMCMCN
o —
0
Q.
C/5
178
NISTNCSTAR
1-3D, V^/TC Investigation
Data Tables and Supplemental Figures
o
o
in in
CM CM
d d
oooooooooooooooooooo
o o o
d d
a
<D
3
C
C
o
£
to
oooooooooooooo
oooooooooooooooooooooo
ouoooooooooooooooooooo
C/3
0)
0)
*j
c
E
3
O
o
TO
11
._
O C
£
0)
i5i?LLLLXIXXLJ.Li.IIIX-5™XXIIII
0)
<D
n:
TO
TO
n:
TO
TO
TO
TO
TO
TO
TO
TO
TO
TO
r^cO'-ooinncsico-^oooO'^Ln'^coTCNicnoo
a>
E
'C
<D
^-CN<M'--r-'<--.-CMCM'<--^'>-'>-^T-'-'>-'<-
Q.
O
(0
™
TO
»
5
0)
M
C
a)
-cf
3
0)
o
oco-q-r-h-csjinSoi'q-m — cot^iDUDor^tnt^P;?^
oooocooocncooo^coooa»a>050>ooooa>05(35cn^^
o
c\i
csi
0)
o
m
CO
c o
re
in
cp cp
o
111
UJ
Ljj
O
CO CO
CD
CO
CO CO 00
CO CO CO CN
CO r- IT) to
^ —
s?
x:
<U
n
t~~
d
0)
Q.
E
O
rj-
CN
tn CO
n n
c\j
>a-
HI
CO
r)
00
C3
in r~
UJ 05 CO
CO C7) CM CO
CO
CO
o
:
oocor^ooN-h^h-cococo
UJ UJ
00 h~
<^
CN to
o
d
CO in
ih CO
Tl-
S
(O
CO CO
O
^
O
to (D
CD CD
CM r—
CM 00 CO
00 CO G> CO CO 00
ir>
d
o d
CNi
'S-
cp CD
'
^OCOOCOCOCOCOCNJCM
mOC35UDCDCDCT>untOCM
05 a> in
d
"
CO CO
o
CO in
o
in
f-
ScMdcbcNdcOTtC^^CMCO'>-tMCM
CO
TO
E
o
o
oi
as
SI
^
—o
<i>
in to UJ UJ
C
i_
O)
c
o
—c
°'-"coco
rMtO'^fOco
_
^
CO
tNco-^"''-"'"cMco'"
-I -I 1-
m
fsw
H
I-
I
<
I m
o P
— m
I
8
inininLnLnLntoinininininininininininininLnLn
c
(D
O
D.
C/5
C
>.
NISTNCSTAR
1-3D.
WTC
Investigation
179
Appendix A
ooooooooooooo
oooooooo
'S-
_]
X
"ai
"oi
13
CM
(U
CD
Q>
O
<U
_J_I"0"0"DT3X)X!"D'0H"D
CD
E
LT)
TO
CL Q.
03
cccccccccocw
CO
CO
to
„-
C\3
C1Q.Q.Q.Q.Q.Q.Q.D-
Q-Cl-
ooooooooooooo
HI-l-l-t-K-l-l-l-l-l-t-la: (T cr tr CH CC IT rr cr
CO CO CO
JZ
q: a: tr
CO CO w
w w w w
I I I I I I X I
to
to
to
to
to
to
to
CO
I I I
CO
to
to
to
to
LL LL LL LL LL LL LL LL LL LL LL LL LL
<
O
o
5
to
O
O
CD CO
CM
en tn
LO
q3
o
"
1^
ro
Uj
o
^'-cMcncDtn-"— tDCNCNtricNjTr
i^r^t^t^r-^cDoocot^cocooooo
I^COCT)'^C35CO'3-'a-COCNJCO
c o
£
LiJLiJLijroro^SS^LUcDr:^^
0)
cJ>CNjcooocot^'>-OcorooooMcNjcoTfincoroincorocncDno
cococotocot^r^i^cDCDr^t^i^^
"Sou.
CO
CO
IN
CM
o
t^
o
—
ID CN
T-^ciiriccigggcNi O
CT)CT)CJ)0>CJ)~~~CJ5 o o o o
c/)
t
to
-£
-S U_
o c
=
0)
^
^
CD
_Q
°°
CM CN
CM
O
'd-
l-HHHI-l-l--J-i-l-l-l
COtX)CX3COOOOOOOtX5t»CXDOOC»CO
(3
2
0
c
CD
a>
CD
ooooooooooooo
oooooooo
cococococococooo
CMINCMCMCMCMCMCNCMrljtNrvJCN
O
^
O
Li.
—
-D
Q. Q)
CO >.
0)
180
NIST NCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
CNini^r~-r^inLnc\ir\jininincNCNinir)incNc\iinLninir)ir)LnLn
CN<N^^^-^^c^JCNC\lCNCNCNC^JC^lCNC\lr^^c^JC^J(^JC^lc^J(^Jc^JC^JC^ic^J
CS|-r^T--r-T-T-r-CSJtN-^'<--^CNrv|T-T-T-(-MCMT-'<-^,-T--^.^
oooooooooooooooooooooooo
O O O O O O O o o o o o
O
d d o
H
r
CO
E
JO
CO
TO
CNI
CNI
CNI
< < < <
o
6 CD CD
CO CO 00 CO
O O o o
a>
c
E
CNJ
CNJ
CNJ
CO CD
:^
h- 1- 1- 1- 1-
m
z
1-
cr cr cc cc
q:
w w w w w
X I I I I I
O
u
£
o
o
o
CNJ
CNJ
CNJ
CNJ
5
CD
CNJ
CNI
CNJ
CNJ
CN
CNJ
CNJ
CNJ
CNJ
m m m m
CNI
CO
hCD
LU
§ 5 §
O O O O O O o oO
z z z
H H hCNI
CNI
m
§
CD ID CD CD CD ID CD CO CD
CO CO CO CD CD CD CD CD CD CD CD CO CD CD CD CD CD CD
CD ca CO CD CD
CM
a: ct
"5
TO "S
TO "(5
LL LL LL LL LL LL
<
o
CNI
LO LD in Ln CD ID in ID CD ID CD CD
Z
_3
I- I-
_l _l
CNI
CNI CO
CNI CNJ CNJ
_j _j _j _j _l _l _l _J _j _1 _1 _l DD CD _l _i _l _J
Li. LL LL LL LL LL LL LL LL L. LJ- LL LU LLI
00
CD CD CD CO
CNI CNJ CNJ CNJ CNJ CNJ CNJ CNJ C\J CNJ CNJ CNI
LU LU LLI LU LU LU
CNI
Q.
o o o o o o
d d
ce a: tr
CO
w w
X X X
ro
O
CO
CD CNI r-~
CD CO in CD ID LD
z
Z z
1- 1- 1-
"5
tr cc ce
q: tr
CD
w w w
X X X
ro
TO
LL LL LL
CN CN LO
in in
CJ5
m
cr cr a:
01
w W w V) w CO
X X X X X X X
ro
TO
LL LL LL
cr
d:
q: cr
"5
TO
ro
TO
TO
LL LL LL LL LL LL LL
O
ID
CD
CO
(J)
(Jl
C35
O
CD
t-^
CNI
in CN
CM CO
CO
10
3
(0
(0
<0
o
*^
.TOiocor^iDcocj)CJ)CNiCNiCNir^r^(0
in CO
CO
-'
™ Uj
CO CO CO CO
CNI
CO CO CO CO CO
'
CNI
CO
TCO CO
CNI
o
o
"eo
c
^'
0)
CL
5 n
in CD CO
CD C3) CNJ
ID in in
CD 00 CD CD
ID CD CD CO <a- •>f >^-
CO
J=
5
O
ID T- r- cn
in
CO
O
O
>-
NCD iri ch
CD CD CD CD
CD
CNlT}-c0-*(32OCDLn
^cor^iD'^j-cococj)
CDIDID-^'^IOIOID
CO
CO
CDCO0OCO->-CO->-COlD
in
O
o
CD CO 00
CO CO
d
d
CO ^'
CD CD CD
!E
CO CO
CD CD CD CD
Tj-
c o
LU
3
CNJ
ID
(D
<3)
CN
O
Tl-
UJ UJ UJ UJ
o
CN
o o
CO CO
o o
o
^
^ o
CN CN
CD CD CD
UJ UJ UJ
CD
CNI
CO
l~^
r-~
CO CO
o
OJ UJ
CO ^aCN CO
^
S
UJ cu
CM
CN CO
'S-
^
'^"CDI^CDCOOCO'^
CO
CO
CN 00
CO in
CO
r~- CD Ol CO 1^ CO 00 CD
CN CN CO -"d-"d-
CD CO 00 CD
OlDT-CNIOCNIt^CO
CJ5
C35
cdcvj'a-cocjicricN-*
(0
0)
Q.
CD in CO TID (35 00 CO
CN CO CO
E
o
CO
"
5
>a-CJ)OCNCOCOC7lOOCNCO
(J)
r- CD
CN CN CN CNI
CO
CNI
CNI
-"a-
'i-
CVJ
o
^
0)
£
E
o
o
01
^ — ^
c o
i;
CO
I
<
2
CO co
in 00
cri 00
CO CO CO
CO CO
CD
in in
CD CD
UJ LU LU UJ
UJ LU UJ
in in
CO
d
CD CD
LL
o
o
o o
<5
n:
-c
o ~
c
—
CD
q
CNI
'•
Ol
-o
c ,_
E o
CD
CO
CO
CO
—= =o
>
QJ
-J
m
_] _i _l _l
(J
CO CO CO CO CO
CD
CO
CO
CO
T3-
CD
CO
CO
CO
•4
CD
CO
CO
CO
o o o o o o
O o o
in CD in in
CNI
d
in
t-
o
CO
_
CO TCN CD CO
•rf
CO CD
CO -"dCN CO
cci
_i _i _i
<y>
CD CD
TO
CO CO op
oooooooo o o 4o
CD
CJ)
cn
o)
o
0)
CO CO CD CD
CO CO CO CO
-ii
Q.
NISTNCSTAR
CO CD CD
CO CO CO
0
1-3D.
WTC
Investigation
181
Appendix A
CM
rsi
CM CM in
CM
CNJ
CM CM
in
OJ
in in in
CM
in in
CM CN
CM
O o O O o o o o O o o o O o o o O
m
m
o O o o o O o o o o o o
O o d o o
d CD O CD CD o (D o o o o o o
in
in in in in
in in in in in in
LD in cn in CM CM CM CM CM CM CM CN CM CM CM CM CM
CM
CD CD CD CD CD CD CD
"O
0)
3
^
n
0
CM
CM
CM
h_J _l L3 _j _] _J _l _j
CM
CM
CM
C\J
CM
CM
ui in in in
FL FL FL FL FL FL
CM CO
CM
hCO _j _J _J
h- CD CD CD CD CD CD
CM UJ LU LU LU UJ UJ
FL
§ 5 §
I I X X
o OO O a U O O O O O I I X X
z
Z Z Z Z Z z z z z
1- H H H H H 1- 1- H 1- 1I X I I
c
o
LO in in in in in in in in in in
CD CD CD CD CD CD CD CD CD CD CD
w
o
§ §
X X
X X
z
z
1- H
a>
'4-1
(A
crircccrDcctcCxxxxxxxxxxxxx
C
E
0)
_CD
Q.
LL
CO
O
o
£
o
u
v.
o
LL
LL LL
LL LL
<
o
I
TO
3
o
0)
0)
*"
LU
rocMrooocMcocMa5a5in-<3-ooc»CMr^cDCDCDCMO
<a-'*-*rO'a-'d--*corO'*M-rorOM-CMCMCMCMcoco
o
CD
(A
0>
J
in
CO CO CD in
00 CD in
C\J
o
CMincD-*Cn'*CDC:5^CT)
in in CM
in
0)
CD
CD
CD
c
a>
<*-•
o
CD
2
o o o o o
in d CO CD
CD
CD CD CO
CD CO
iri
iri
CD in CO (35
CD CO
CD 00
d
CD CD
O
<- CD CM
CD
CO CD T-^ CM in
CO
O o o
c o
CD
o o CO
o
CD
^
UJ LU LU LU LU LU UJ
CT) 1^ CO
co CO CO CO CO
CM
TO
!E
O
CO
in CD
00
~5 CO
CO^LUTOp^CN^^
cmcoS'°cm"-'cmco
^t— >3-'<a-cocDr^cDr-~f~-ocM-*t^cr)CMooci>coin
3
u
^-cDCDCMCJ>incDi^cj)0>i^coocoincoooo'*c35
CMCMCMcocMCMCMCMcoco-^co-^-^co-^in-^inin
0)
a
E
>
a>
E
o
o
O
i;
C O
CO
I
<
£
o c
U)
g
Q. "o
cn
5 o
.Q
S
o
-I
2
CD
CD
5
m
o
Jj^cocooin^'-h-T-cDCDCMTj-cDcDocrir^cncqco
h~o6t~-CNir^cbdi^^-o6iricddcMCJ>codc6cD
coro-^'^corO'^inincDinint^incDr^i^r-- 00
LULUUJujujujujuj'-Sfein^^LUOO^ScM
^
CD
-
182
^
LL
oocooooooooocM-"—
-I -1 -J I<D
o o o o o
-
0 >
<D
5
^ E
0)
_ 5
CDCDCDCDCDCDCDCDCDCDCDCDCDCDCD;?;S;i;5;5;
cooocococcoooDcocpcooooocoapoogJg^g^g^g^
oooooooooooooooSSSSS
cj5Cncpcpcnc3>cpcnc35Cpcj)a)CDcj5CDj^JoJ^J^^
o
NISTNCSTAR
1-3D,
WTC
Investigation
311
5
1
Data Tables and Supplemental Figures
A.3
DATA TABLES FOR IMPACT PROPERTIES
The following
table
summarizes the Charpy
test results.
Table A-15. Charpy V-notch impaci data for samples
t
Specimen
Test
Building
Thickness
Specimen
Temp
Location
in.
Orientation
F
1-42-97 100
0.2
T
77
1-142-97 100
0.2
T
1-142-9" 100
0.2
T
1-142-97 100
0.2
T
1-142-97 100
0.2
T
32
1-142-97 100
0.2
T
32
1-142-97 100
0.2
T
-4
1-142-97 100
0.2
T
-4
1-142-97 100
0.2
1-142-97 100
0.2
1-142-97 100
tesited."
Energy-
Sp^ci men
Ent'rg>'
Corrected
ft-Ibf
ft-Ibf
ID
14.2
25.6
SL-1
77
12.8
23,0
SL-2
N8-C1 Bl Spandrel T
77
13.3
24.0
SL-3
N8-ClBlSpaiidrcl_T
14.0
25.2
SL-7
N8-C1 Bl Spandrel T
14.9
26.9
SL-8
N8-CIBlSpandrel_T
13.4
24.1
SL-9
N8-ClBlSpandrcl_T
Specimen Type
N8-C1 Bl Spandrel T
1.8
21.2
SL-10
N8-ClBlSpandrcl_T
13.3
23.9
SL-1
N8-ClBlSpandrcl_T
T
-4
15.2
27.4
SL-12
N8-C1 Bl Spandrel T
T
-40
12.6
22.8
SL-1
N8-C1 Bl Spandrel T
0.2
T
-40
12.6
22.8
SL-14
N8-ClBlSpandrel_T
1-142-97/100
0.2
T
-40
1
1.5
20.7
SL-1
N8-ClBlSpandrcl_T
1-142-97 100
0.2
T
-76
12.2
21.9
SL-4
N8-ClBlSpandrcl_T
1-142-97 100
0.2
T
12
2.4
4.4
SL-5
N8-ClBlSpandrcl_T
1-142-97100
0.2
L
77
43.3
78.0
ST-1
N8-C1B1-Spandrcl L
1-142-97 100
0.2
L
77
42.7
76.9
ST-2
N8-ClBl-Spandrcl_L
1-142-97 100
0.2
L
77
39.3
70.7
ST-3
N8-C1B1-Spandrel L
1-142-97 100
0.2
L
32
39.4
70.8
ST-7
N8-C1B1-Spandrel L
1-142-97'100
0.2
L
32
38.6
69.4
ST-8
N8-CIBl-Spandrcl_L
1-142-97 100
0.2
L
32
39.6
71.3
ST-9
N8-C1B1-Spandrel L
-1
1
1-142-97 100
0.2
L
-4
43.0
77.5
ST- 10
N8-C1B1-Spandrcl L
1-142-97 100
0.2
L
-4
42.5
76.6
ST-1
N8-C1B1-Spandrcl L
1-142-97 100
0.2
L
-4
40.0
72.0
ST-12
N8-ClBl-Spandrcl_L
1-142-97 100
0.2
L
-40
32.5
58.5
ST- 13
N8-ClBl-Spandrel_L
1-142-97 100
0.2
L
-40
39.5
71.2
ST- 14
N8-C1B1-Spandrel L
1-142-97 100
0.2
L
-40
40.1
72.2
ST- 15
N8-ClBl-Spandrcl_L
42-97 100
0.2
L
-76
15.3
27.5
ST-4
N8-ClBl-Spandrel_L
1-142-97 100
0.2
L
12
9.8
17.6
ST-5
N8-ClBl-Spandrel_L
1-142-97 100
0.2
L
77
35.3
63.6
NL-1
N8-ClMl-Flangc_L
1-142-97' 100
0.2
L
77
41.0
73.9
NL-2
N8-C1 Ml -Flange L
1-142-97'100
0.2
L
77
38.7
69.6
NL-3
N8-ClMl-Flange_L
1-142-97 100
0.2
L
32
37.4
67.4
NL-7
N8-ClMl-Flangc_L
1-142-97/100
0.2
L
32
38.1
68.6
NL-8
N8-ClMl-Flangc_L
1-142-97 100
0.2
L
32
39.2
70.5
NL-9
N8-ClMl-Flangc_L
-4
38.2
68.8
NL-10
N8-ClMl-Flange_L
1-1
-1
1-142-97 100
0.2
L
1-142-97 100
0.2
1
-4
38.0
68.5
NL-1
N8-ClMl-Flangc_L
1-142-97 100
0.2
L
-4
37.3
67.2
NL-12
N8-ClMl-Flangc_L
1-142-97 '100
0.2
1
-40
35.2
63.4
NL-13
N8-C1M
1-142-97 100
0.2
L
-40
32.5
58.5
NL-14
N8-ClMl-Flange_L
NISTNCSTAR
1-3D,
WTC
Investigation
1
-Flange
1,
183
Appendix A
Test
Building
Thickness
Specimen
Temp
Location
in.
Orientation
F
1-142-97/100
0.2
L
-40
1-142-97/100
0.2
1
1-142-97/100
0.2
L
1-142-97/100
0.2
1-142-97/100
1-142-97/100
Energ\'
Energy
Corrected
ft-lbf
it-lbf
Specimen
ID
34.1
61.3
NL-15
-76
25.5
46.0
NL-4
N8-C Ml -Flange
1,
12
15.2
27.3
NL-5
N8-C1 Ml -Flange
L
T
77
10.4
18.7
NT-1
N8-C1M1-Flangc T
0.2
T
77
10.9
19.7
NT-2
N8-C1M1-Flangc T
0.2
T
77
10.9
19.7
NT-3
N8-C1M
1-142-97/100
0.2
T
32
10.6
19.0
NT-7
N8-C1 Ml -Flange T
1-142-97/100
0.2
T
10.9
19.6
NT-8
N8-C1 Ml -Flange T
1-142-97/100
0.2
T
32
10.1
18.2
NT-9
N8-C1M1-Flange T
1-142-97/100
0.2
1
-4
10.2
18.3
NT-10
N8-C1 Ml -Flange T
1-142-97/100
0.2
1-142-97/100
0.2
1-142-97/100
-1
Specimen Type
N8-ClMl-Flangc_L
1
1
T
-Flange
T
-4
9.5
17,1
NT-1
N8-ClMl-Flange_T
T
-4
10.8
19.4
NT- 12
N8-C1M1-Flange T
0,2
T
-40
9.9
17,8
NT-1
N8-C1M1-Flange T
1-142-97/100
0.2
T
-40
10.5
18.9
NT- 14
N8-C Ml -Flange T
1-142-97/100
0.2
T
-40
10.5
18.9
NT- 15
N8-C1M1-Flangc T
1-142-97/100
0.2
-76
10.5
18.9
NT-4
N8-C1M1-Flangc T
1-142-97/100
0.2
T
-112
5.6
10.1
NT-5
N8-C1 Ml -Flange T
1-142-97/100
0.2
L
77
44.1
79.4
HL
N8-C1M1-HAZ
L
1-142-97/100
0.2
L
77
37.4
67.3
HL
N8-C1M1-HAZ
1,
1-142-97/100
0.2
L
77
49.8
89.7
HL
N8-C1M1-HAZ_L
1-142-97/100
0.2
L
32
32.2
57.9
HL
N8-C1M1-HAZ_L
1-142-97/100
0.2
L
32
40.2
72.4
HL
N8-C1M1-HAZ_L
1-142-97/100
0.2
L
32
44.2
79.5
HL
N8-C1M1-HAZ
1-142-97/100
0.2
L
-4
22.6
40.7
HL
N8-C1M1-HAZ_L
1-142-97/100
0.2
L
-4
42.8
77.0
HL
N8-C1M1-HAZ_L
1-142-97/100
0.2
L
-4
23.6
42.6
HL
N8-C1M1-HAZ_L
1-142-97/100
0.2
T
77
9.8
17.7
HT
N8-ClMl-Weld_T
1-142-97/100
0.2
T
77
9.9
17.8
HT
N8-ClMl-Weld_T
1-142-97/100
0.2
T
77
9.1
16.3
HT
N8-ClMl-Weid_T
1-142-97/100
0.2
T
32
9.2
16.6
HT
N8-ClMl-Wcld_T
1-142-97/100
0.2
T
32
9.4
16.9
HT
N8-ClMl-Weld_T
i
1
L
1-142-97/100
0.2
T
32
9.2
16.6
HT
N8-ClMl-Weld_T
1-142-97/100
0.2
T
-4
9.8
17.6
HT
N8-ClMl-Weld_T
1-142-97/100
0.2
T
-4
9.6
17.3
HT
N8-ClMl-Wcld_T
1-142-97/100
0.2
T
-4
8.9
15.9
HT
N8-ClMl-Wcld_T
1-451-85/88
0.2
L
77
43.2
77.8
CIO L
ClO-ClMl-FlangcL
1-451-85/88
0.2
L
77
47.5
85.5
CIO L
C10-C]Ml-Flange_L
1-451-85/88
0.2
L
77
47.2
85.0
CIO L
ClO-ClMl- Flange L
1-451-85/88
0.2
L
32
47.2
85.0
ClOL
ClO-ClMl - Flange L
1-451-85/88
0.2
L
32
46.5
83.7
ClOL
ClO-ClMl-FlangcL
1-451-85/88
0.2
L
32
51.5
92.7
CIO L
ClO-ClMl - Flangc L
1-451-85/88
0.2
L
-4
47.1
84.8
CIO L
ClO-ClMl - Flange L
1-451-85/88
0.2
L
-4
48.3
86.9
ClOL
ClO-ClMl-FlangcL
l-451_X5/88
0,2
L
-4
47
85,5
ClOL
C10-ClMl-Flange_L
184
5
NISTNCSTAR
1-3D,
WTC Investigation
Data Tables and Supplemental Figures
Specimen
E n cr*'^
Test
Temp
Building
Thickness
Specimen
Location
in.
Orientation
1-451-85 88
0.2
L
-40
1-451-85 88
0.2
L
-40
1-451-85 88
0.2
L
Energy
Corrected
ft-lbf
ft-lbf
Specimen
ID
46.3
83,3
CIO L
C 0-C
51.4
92,5
ClOL
ClO-ClMl-FlangcL
-40
46.5
83,7
CIO L
ClO-ClMl-FlangcL
-76
43.7
78,7
ClOL
ClO-Cl Ml
12
44.8
80.6
CIO L
ClO-ClMl- Flangc L
op
Specimen Type
1
1
M
-
1
Flange L
-45 -85 88
0.2
L
1-451-85 88
0.2
L
1-451-85 88
0.2
T
77
1
1.2
20.2
ClOT
ClO-ClMl -
1^51-85 88
0.2
T
77
12.4
22.3
CIO T
ClO-ClMl- Flange T
M51-85
88
0.2
T
77
1
1.4
20.5
ClOT
ClO-ClMl -
Flange
T
1-451-85 88
0.2
T
32
12.4
ClOT
ClO-ClMl -
Flange
T
M51-85
88
0.2
T
32
1
1.2
20.2
ClOT
ClO-ClMl- Flangc T
1-451-85 88
0.2
T
32
1
1-451-85 88
0.2
T
-4
1-451-85 88
0.2
T
-4
1-451-85 88
0.2
T
-4
M5 1-85
88
0.2
T
-40
1_;51.85 88
0.2
T
^0
M51-85
0.2
T
-40
1
-76
1
1
M51-85
-1
-
Flange L
Flange
T
1.6
20.9
CIO T
ClO-Cl Ml
11.8
21.2
ClOT
ClO-ClMl - Flange T
1
1.4
20.5
CIO T
ClO-ClMl - Flange T
1
1.0
19.8
CIO T
ClO-Cl Ml
-
Flange
T
10.7
19.3
ClOT
ClO-ClMl -
Flange
T
10.9
19.6
ClOT
ClO-ClMl -
1.3
20.3
CIO T
C 0-C
1.3
20.3
19.3
ClOT
ClOT
ClO-Cl Ml
10.7
-
Flange
T
Flangc
T
-
Flange
T
-
Flange
T
ClO-ClMl -
Flangc
T
1
1
M
1
88
0.2
T
1-45 1-85 '88
0.2
T
1_451.S5 88
0.2
T
77
11.1
20.0
CIO
HAZ
ClO-ClMl-HAZ T
1-451-85 88
0.2
T
77
12.0
21.6
ClO-ClMl-HAZT
1-451-85 88
0.2
T
32
1
1.4
20.5
M51-85'88
0.2
T
1
1.2
20.2
ClOHAZ
ClOHAZ
ClOHAZ
0.2
T
32
1
1.0
19.8
CIO
HAZ
ClO-ClMl-HAZ T
1-451-85 88
0.2
T
-4
12.3
22.1
0.2
T
-4
1
1.7
21.1
1-451-85 88
0.2
T
-4
1
1.3
20.3
1-451-85 88
0.2
T
-40
10.3
18.5
ClOHAZ
CIO HAZ
CIO HAZ
ClOHAZ
ClO-ClMl-HAZT
1-451-85 88
1-451-85 88
0.2
T
-40
1
1,0
19.8
ClOHAZ
ClO-ClMl-HAZT
1-451-85 88
0.2
T
-40
9.9
17.8
CIO
HAZ
ClO-CIMl-HAZ T
1-451-85 88
0.2
T
-76
10,5
18.9
ClOHAZ
ClO-ClMl-HAZT
1-451-85 88
0.2
T
12
9.2
16.6
CIO
HAZ
ClO-ClMl-HAZT
1-603-92/95
0.4
T
172
48.9
48.9
C80_T
C80-wcb-column 603_T
1-603-92 '95
0.4
T
172
45.0
45,0
C80 T
C80-web-colunin 603_T
1-603-92 '95
0.4
T
140
45,4
45,4
C80_T
C80-wcb-column 603_T
1-603-92 '95
0.4
T
140
46,3
46,3
C80_T
C80-web-column 603_T
1-603-92/95
0.4
T
77
28,9
28,9
C80_T
C80-wcb-column 603_T
1-603-92/95
0,4
T
77
24,4
24,4
C80 T
C80-wcb-column 603_T
l-603-92'95
0.4
T
77
25,1
25,1
C80_T
C80-wcb-column 603_T
1-603-92/95
0.4
T
32
7,1
7,1
C80_T
C80-wcb-column 603_T
1-603-92 '95
0.4
T
32
7,3
7,3
C80_T
C80-wcb-column 603_T
1-603-92 95
0.4
T
32
13,2
13,2
C80_T
C80-web-column 603_T
1-603-92 '95
0.4
T
-4
3,8
3,8
C80_T
C80-web-column 603_T
0.4
T
-4
4.0
C80 T
C80-wcb-column 603_T
1-603-92 '95
NISTNCSTAR
1-3D.
WTC
Investigation
-1
-1
12
1
4,0
ClO-ClMl-HAZT
ClO-ClMl-HAZ T
ClO-ClMl-HAZT
ClO-ClMl-HAZ T
ClO-ClMl-HAZT
185
Appendix A
^npri men
Test
Ener^'v
Building
Thickness
Specimen
Temp
Energy
Corrected
Location
in.
Orientation
"F
ft-lbf
ft-lbf
1-603-92/95
0.4
T
-4
3.6
1-603-92/95
0,4
L
172
1
1-603-92/95
0.4
L
172
1-603-92/95
0.4
L
140
Specimen
ID
Specimen Type
3.6
C80 T
C80-web-column 603_T
0.5
C80_L
C80-web-column 603 L
121.1
121.1
C80_L
C80-wcb-column 603 L
115.1
1
15.1
C80_L
C80-wcb-column 603_L
10.5
1
1
1-603-92/95
0.4
L
140
103.9
103.9
C80 L
C80-wcb-colunin 603_L
1-603-92/95
0.4
L
77
45.6
45.6
C80 L
C80-wcb-column 603 L
1-603-92/95
0.4
L
77
47.4
47.4
C80_L
C80-wcb-column 603_L
1-603-92/95
0.4
L
77
68.3
68.3
C80 L
C80-wcb-column 603_L
1
-603-92/95
0.4
L
32
36.3
36.3
C80_L
C80-wcb-column 603 L
1
-603-92/95
0.4
L
32
44.1
44.1
C80_L
C80-wcb-column 603_L
1-603-92/95
0.4
L
32
35.6
35,6
C80_L
C80-wcb-column 603_L
1-603-92/95
0.4
L
-4
8.9
8.9
C80_L
C80-web-column 603_L
1-603-92/95
0.4
L
-4
7,0
7.0
C80_L
C80-wcb-column 603_L
1-603-92/95
0.4
L
-4
3.8
3.8
C80 L
C80-web-column 603_L
unknown
0.2
L
77
29.1
52.3
Tl-1
Angle (2X1.5X0.25 inch)
unknown
0,2
L
77
33,5
60.4
Tl-2
Angle (2X1.5X0.25 inch)
unknown
0.2
L
77
30.7
55.3
Tl-3
Angle (2X1.5X0.25 nich)
unknown
0.2
L
32
18.2
32.7
Tl-7
Angle (2X1.5X0.25 inch)
unknown
0.2
L
32
27.5
49.4
Tl-8
Angle (2X1.5X0.25 inch)
unknown
0.2
L
32
19.0
34,2
Tl-9
Angle (2X1.5X0.25 inch)
unknown
0.2
L
-4
8.3
14.9
Tl-10
Angle (2X1.5X0.25 inch)
unknown
0.2
L
-4
9.8
17.7
Tl-1
Angle (2X1.5X0.25 inch)
unknown
0.2
L
-4
8.6
15.5
Tl-12
Angle (2X1.5X0.25 inch)
unknown
0.2
L
-40
5.5
9.9
Tl-13
Angle (2X1.5X0.25 inch)
unknown
0.2
L
-40
1.9
3.4
Tl-14
Angle (2X1.5X0.25 inch)
unknown
0.2
L
-40
1.6
2.9
Tl-15
Angle (2X1.5X0.25 inch)
unknown
0.2
L
-76
1,6
3.0
Tl-4
Angle (2X1.5X0.25 inch)
unknown
0.4
L
172
100,3
100.3
TR
unknown
0.4
L
172
1
13,8
13.8
unknown
0.4
L
140
101.2
101.2
unknown
0.4
L
140
96.9
unknown
0.4
L
77
unknown
0.4
L
77
unknown
0.4
L
77
unknown
0.4
L
32
unknown
0.4
L
unknown
0.4
L
unknown
0.4
unknown
unknown
186
Rod
(0.92 inch)
TR
Rod
(0.92 inch)
TR
Rod
(0.92 inch)
96.9
TR
Rod
(0.92 inch)
22,4
22.4
TR
Rod
(0.92 inch)
41.7
41.7
TR
Rod
(0.92 inch)
33.1
33.1
TR
Rod
(0.92 inch).
21
.3
21.3
TR
Rod
(0.92 inch)
32
19.8
19.8
TR
Rod
(0.92 inch)
32
7.5
7.5
TR
Rod
(0.92 inch)
L
-4
5.4
5.4
TR
Rod
(0.92 inch)
0.4
L
-4
3.2
3.2
TR
Rod
(0.92 inch)
0.4
L
-4
6.1
6.1
TR
Rod
(0.92 inch)
0.4
T
172
27.4
27.4
N13
Scat
0.4
T
172
26.9
26.9
N13
Scat
0.4
T
140
21.6
21.6
N13
Scat
0.4
T
140
23,0
23.0
N13
Scat
1
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
Specimen
a.
Test
Building
Thickness
Specimen
Temp
Location
in.
Orientation
F
0.4
T
77
6.1
6.1
N13
Scat
0.4
T
77
10.3
10.3
N13
Scat
0.4
T
77
6.7
6.7
N13
Scat
0.4
T
3.6
3.6
N13
Scat
0.4
T
3.0
3.0
N13
Scat
0.4
T
3.3
3.3
N13
Scat
1-142-97 100
0.4
T
172
36.8
36.8
N8
Scat
1-142-97 100
0.4
T
140
33.2
33.2
N8
Scat
1-142-97 100
0.4
T
140
33.2
33.2
N8
Scat
1-142-97, 100
0.4
T
104
26.3
26.3
N8
Scat
1-142-97 100
0.4
T
104
28.7
28.7
N8
Scat
1-142-97 100
0.4
T
77
20.7
20,7
N8
Scat
1-142-97 100
0.4
T
77
22.4
22.4
N8
Scat
1-142-97 100
0.4
T
77
20.6
20.6
N8
Scat
1-142-97 100
0.4
T
-40
1.9
1.9
N8
Scat
'^")
Energy
Corrected
ft-lbf
ft-lbl
Specimen
ID
Specimen Type
1-142-97 100
0.4
T
-40
2.9
2.9
N8
Scat
unknov\n
0.4
T
172
34.2
34.2
M4
Scat
unknov\n
0.4
T
140
28.7
28.7
M4
Scat
unknown
0.4
T
140
28.3
28.3
1VI4
Scat
unknown
0.4
T
104
19.9
19.9
M4
Scat
unknown
0.4
T
104
21.1
21.1
M4
Scat
unknown
0.4
T
77
10.8
10.8
M4
Scat
unknown
0.4
T
77
12.9
12.9
M4
Scat
unknovv n
0.4
T
77
11.9
1
1.9
M4
Scat
unknown
0.4
T
-40
2.1
2.1
M4
Scat
unknown
0.4
T
-40
2.5
2.5
M4
The
oinentation of spandrel specimens appears to be reversed because the specimen identifications (L
Scat
and T) were
in
reference to
the building, rather than the rolhng direction of the steel.
NISTNCSTAR
1-3D.
WTC
Investigation
187
Appendix A
A.4
DATA TABLES FOR ELEVATED-TEMPERATURE PROPERTIES
The following
188
tables
summarize the data from the elevated temperature
tensile tests.
NISTNCSTAR
1-3D,
WTC Investigation
Data Tables and Supplemental Figures
c\j
CO
^
Ln CD
2
CM
O
o
O
2
CN
o
o
O
2
CM
o
o
O
2
2
CM CM CM
O (J O O o o o o u o ooooouoooo
o O o CO CO CO CO CO 00 CO ocDOOocioocio
OOO zZzzZZ
T--cMor)-<a-Lnopci5-^CDr»^
<<<<<<< LLLLLLU-LLLLILLLLLLL
Q.
E
2
O
o
O
01
CSI
CO
CO
CO
oo
2
-5 -!S -5
"c "c
-5
^^^
c c c c
°^^
-5
"c
c c c c c c c
J2
*
fOcncOO-i-COCDincOOCOOO^r^CDOM-CMCDCXlOCDlO
(A
lDCDm<D'q-r>-COCDCDCDt^COCOCOCDCD<DCDr-~t^h-0Or^t-^
C
£
3
O
u
o
CC
LU
cocDt~-.CMCO'53-Lno5CT)CDr-uncO'g-h-coM-coLDOCT>h-r--'3-
COCOCMCDCM'^incOCOrOn^'^mCMCMCMCMCMCOCMCMCO'^
Q)
CN-^-q-cDLnr-inr-^incDO-g-roocO'^cMcococotDiDCT)
E
t^h-'^CDOCOCNCMCO-^O'^r^CMCDinO'— CDOCJ^OinCT)
'iZ
03
<U
«
0)
ininiD-srcO'^T-cDCDCD'a-cot-T-oococot^uo^-'a-LncoT-
•-5
a
O(NCJ)CDCM-^-<-OC75CDOC0r0C0'^^
00 CO CO
—
CO
IT) 1^
v> CM 05 uj
CO
CM CM
CD CO CD vn
CM CM 06
t~~
CO
^
,
cocoi^<0'VCM'<-ci>c3>a5i^'^cM'-^iJ^^cot^r~.r-.LOCM
CO
3
(A
0>
I-
"CI
o >
m
,
£
CD
^ ^
O
^- C5
LO CO
q- CO CO CM
CJ>
CM
CNJ
r-•>a-
10 CO CD CO
rvj CO
CO CO CM
OOCOLDCM^-^TCDt^OO
CO
(/>
V
o
*->
i^-^r^o-<a--^ooM--<a--<j-cocMLf>r^
O
O
CD
CD
r^COOCOOCNCNJCD
(H
c
CO
0)
5
c
£
3
.£ Qo >Q.
S2
0)
^^^C7)m-^cDiDcoiDcocoococ35incocDCDcoir^tnc35
•^>-<:50->-'^cT>-q--<r-q-ro^M-^oo--ooooooo
cDCDincDCDCDcnr^t^r^r^r— r— r^r-r--r^t~~(^i^r— t^t^r-
a
ooooooooooooooppoooooooo
CD<zj0CD0C3C>C3CDCD0CiC)CDCD0CD0CDCDC30C30
E
0)
<D
O)
UCJOOOOOOCJOOOOOCJOCJUOOOOOO
11
S
Q.
CO
LlLLLLLLLLLLLLLlU-LlLLLLLLLLLLLLLlLLLLLlLLLLLLLL
c
(0
^~oooooooo
Q.
E
flj
„
o
o (-,(-,00000 ir^ir^ooocnLninoin
oo
°Q
oocnfr^i^ooooin
oo
oo
c?
c? cn
mcDCD'^'^CO-<3-LncDCD '^'^co-'a-cDioincncDcD
o
o
D. Q-
,-^^T-,.-^T-0000000
00000000000000
CO0O0OCO0OCX3CO0OCO0O
cooococooocooococooo
£ i
~
m
O
O
Ol
5
^
CO
CDCDCDCDCOCDCDCMCMCMCNr^lCMr^l7ZrZ|7;^j;^[r,loScn
o
^
c o
3 =
O
COCOCOCOCOCOCO^M-Tj-^rJ-Tj-^^I^S^^I^I^Ig^!^^
.2 U.
NISTNCSTAR
3D,
1-.
WTC
Investigation
189
Appendix A
roroa3ran3nj(DnjcDnjajroi2iS-2iSi?pfctfcp
CDlDCDtDCDCDCDcntDtDlDlDoOOOOOOOOOCMOjrMCNCM
T— r-T—
T-T— T— r-r-T— T—
^^r-T— CT>CT>CT)O^CT>
LLU.LLLLLLU.LLLLLLLLU.LLLJ.L.U.LLLLLLU.LLLLLL
-r—
®
ID
Q. cl
>
>
CNCNCMCNJC\iCNCMr\JCNC\ICNC\J-^"*-a'-a--^CMCMCNCNirM
CN
_j
CO
_l _] _J
_l
Ln in en
CNJ CM CM in "
_j _J _j
00
in in
_j _j _J _j
LI. LL Ll_ u. CM CM _1
in un in in in in cb
CD CD CD CD CO CO 00
I ii ii X i
in
in
in
cn
to CD CD CD
OOOO
2 uT
"* CO
CO 00
cr>
o oo
CD CQ
LU
CM to
o
§ §
to a- in
_j -j _j _j
_j _j .j _j
u. LI. u. LL.
CD d o
oo 00 CO 00
OooOOOOoOOO
in 00 CD
-2 CO 00 CJ) -5
~p -5
"c
CD di
CD
TO
^
d
CN CO
CO CO CO CO
I IX X
X XX X
c a
tz
a
CD to C31
CO CO CO
(r>
,
T3 XI "O XJ
p p c c
0)
£
cO'--^ootocnin->— CNtM'-co,oOO'-'3-t^Lnin_,cNjcJiCMcoco
incocO'*oocOLn'*incnrocMjrrO'3-cooinco2<'^'*inotD
T}-T--^Tr^M-'3-^M-cocM'>-'^-^'3-'a-'^cN'<-'*'m'*cocM-^
t^otxiooooocDcninLnoinotooocDcocM'-ocMO
'^
^ 1
o >
2
o
cninT-cj^ocMCMT-T-ocj)^
in'-CNiotointoocMCJ)-^
rn
^
CO
cj)
CO
Q
m
CO
in 00
in
to
CO CM CN
e3>
CM'-tMCMCMCNirMCMCM'-'-
CO
o
>"
cp^~oc^M;lnt^^oLncp^--^lnco*OLnco^^^oolno(5CD
CO^COCOCOCOOOCMCOCMCM'-' cOCOCOCMCM^'^LncOCOCM^
£
CO
to
CL D. CL D. Q. CL
03
> > > > > > ^
2Z
z
CO
tn
CT3
CT3njCD(TJ(D(DCDtD(D
CO
"c "p "p
I
P^^^^^'p'p'p
cocDOCJ>cT>cTJOocooO'^LninLnr^r^un-^'^'^'^r^c7>in'^ao
„
CD
CDCJ>tTJOOCOQOCOOOCOt:nt7)tJ^CJ>COCOCyiCDC7>C3>CDCJlCJ>CJ>CJ)CJ)
CJ) tjl 00
mr\I-rt--rt-^-^T:1--rJ-t;f'^'<:3-'<^'^T3-'<:3-T:t'd"'^^*^CMCMCMCMCM
CM
CJ)
-i^
CD
CD
CD CD CD
oooooooooooooooooo
dc3CDCDC3CDtDtbcbcDC3CD(DCD(DOOCD
OO OO
d d
<u
>
9:
E
°
c
TO
^
-C3
LL
E
I-
X-DXl"D"D"0"D"0"0"DX!'aX3X)'D"a"0"0
O
o
o
in
o
oo
oo
g- in CD CD
o
,
-C3
-C3
-C3
-C3
TO
a
TO
Li.
LL LL LL
oooo^oooo
ooounx,looom
a-incDcD'^-^mcDtD
g « O
C
c £ 0
OJ
^ >
ra
t o
CD
CO
CO
CO
CD CD a> CD
CO
00 00
CO
CO CO 00 00
Tj-
"d-
OS-
0)
CD
CO
CD
CO
CD CD
CO
CO CO
CD
CO
00 00 00 op
CD CD CD CD CD in in in
00 CO 00 00 CO
ro
CM CM CM
CO
CO
00 op op op CO CD CD CD
43CO CO CO
in
O)
CM
CD
CO
ooooo
in
S?
CM 00
00
CD CD CD
CD CD
CO in in in in in
4 4
4 *
j
o
o oo
Ooo oo
o Oooo
oooooooooo
in O)
CD CD CD
CD CO CO CO CO CO CO CO CO CO CO
a> cn CD a> CD
(J>
CJl
CJ)
P
a. 0)
CO >.
190
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
ccccccccccccccccccc
ccccccccc
ID
CNI
in in Lo ir> in LO in in in in in in in in in in in .£ .£ .£ c c c c c
(N (N (M CN CM CN CN (M CM C\i C\JC\|CMrMCMC\ICMCM|^l^l^
1^
(6 di (6 di
CO CO ro CO CO
X X X X X X X X X X
CD
d d
o o o o o o o o o
Q DQ Q Q Q Q D Q
II
II
II
K- I- 1-
II
II
Kl—
II
II
h-l—
II
II
I— I—
E E E E E E E E E
o o o o o o o o o
d d d d d d d d d d d d d m m m
xxxxxxxxdoo
d d d d
xxxxxxxx
§§§§§§x'<x'<xxxxxxxxxxxxxxxxxxxxxxx
X
c c c c c c c c c c c c c c c c c c c
c d d d d d d d
c c c c c c ininininininininminLninmininininmin
CMCMCMCMCMCMCMCVI
II
II
II
II
II
II
U)
t/1
t/1
to
Ul
Q Q Q Q Q D
l/D
CMCMCNJCNJCNJCMCMCMCMCMCM CMCMCNCslCNCNCMCMCOCOCOCOCOCOCOrO
y5uit/)i/)u)ui(/)i/)[n(/)[/>
^O^^OOQ^Qi^QJ^
t/Dt/)i/itnt/)t/5t/ii/5i/)i/)t/)(/5(/>t/){/)in
CU_aj
OJ
CD^_QJ_QJ_aJ
<U
CD
CD
(1)
CD^
(D
OJ
oooooooooooooooccccccccccccccccccccccccccc
Q;Q:ccD^cccc£rQ:Q:Q:Q:Q:c!:Q:a:<<<<<<<<<<<<<<<<<<<<<<<<<<<
"O -O
"O "O
05t/)t/)cflioiotou5c/5U5t/)c/>too5c/)t/)t/:t/)t/)y)u)yD(Owt/)i/)y5LOU5(/5c/)c/)U)wt/)t/ic/>tntni/)c/)t/)
t/)co(Ot;5tnt/)«i/)t/)(/)y5U5wt/)wu5y:int/)t/)u5u:tot/5tO{/)int/)iO(/)c/5c/)i/)ult/)U)t/iu5towt/)(/)
I- I-
K
h-
H
H
I- I-
jn
CO
a
o
K
m in
0
V)
1-;-
K
CO O)
CO CO CO
in in in
0
in
-S -5 -5
"c "c "S
CO
-5
I-
<<<<<^'^^^^COt-CN
CD CM CO
K<<<<<<<<
000 l-l-l-l-l-l-ZZZ 000000000000000000
T- CO
H- H- H- CO CO
in in
in
H
< < < < <
0:0:0:0:
CM CO "J- in
q:
cr CC
1- 1-
KHH
1^ (D CM
9 9 9 9
E
I- I-
^ ^
™
"c "c 'c "£
I— I— I— I— I— I— I— I—
^CM-^COCOCTJ'^CMCO I—
CMCNICMCNCNCMCMCMCMCNCDCDCQCOCDCDCQCQ
cofococococococorocococococococoroco
T-T-r-^r-r-inininininLninLO
<<<<<<<<<
o
^
J.'
^
(0
(0
ro
"c
ro
CD
CD
CD
^ ^^ ^
CD
CD
CD
^ ^ ^ ^
C
CD
"c
CD
CD
CD
^
3
inmr~-rMN-c\jcMcocNOCDa>c3>roini3)o^cDcDc3>'*CMo:P5
COCOCOCOCD'^CMCNJ'^CO'^COCOc'^l^COCDin-'-CMCDCDCDcC
o
T3
^
CD
C
TJ :d :d
^ ^ ^
CD
~c
CO
m
CJ)
"'3
in
»-
10
*^
3
r^'^incDincol^cDOco^^-r- ^cxjoOT-incNioocncMcocNcio =
CMCM'-CNCOCM"'c\JCO00COC0CO
(0
0)
-"a-COCOCOCOT-COCOCOCOCO
(0
cocMr-.cocMr^cJ5'^,^coor--r^
0)
m^cDin^inTT
coh-CMCncMcoinco>i;cocoO'---iS"-0'--^CT!
ininininincocM^"-''^cocM'- c'^inininmcsi
<»
(0
cDOiOh-r-'-coa)
c
o
po o>
CO t--:
O
CD
Csi
CD CO
CO
CD
CD 00 <- CO
in CM
in in in in
CDCOCDCMincOOCMOCDCDI^CNLncOOO
Tj-Csl'^^CDCOCDOtD-'— l-^CMCOOOOCO
inincocM-^-^f-cNininm-^cocMT-T-
CM in CD CD
cri CD
CD
1^
1^
cD in CO 00 CD
CD
CO
in CO CM
d
r\i
T-
co
CO
d
COCOCOCOCDCD^I^COCJ)
h-dT^inro^c6dcbcD
CMCMOOr^OOCO^CslCMT-
d d
0)"
3
2
ra
>-
COCMT-h~CDCOinCM"^incOCO_,_,_,0'^Tl-OCD
OincDC0CDCDT-c0(D<0'^'''m^lS'^'^'^"'f^
ini^cDincDcocDinr^cDcocococxjcoi^cNjocDCMin
00CX)f^CDCOCDCDCDC0incOCOincOLncDCX>CX)C0CX5Csl
COCOCOCOrrcOCOCMT-T-CO^T-TrcOCOCMCNT-T-i-
^in-^rj-inincMT-^ <5;cooor>-.^g3h-inc\ih--^
cdincNcor^cNT-'d^
rodcodrr^cood-^dco
CD'-CDCOCOO^GO^COCOCMN-inCDCMCDCDOCO-'—
rJ-TtcO'q^COCMCM'-mtM<N->-"''^^COCOrOCOC\J
£
0)
a
E
iricD-^'r-^cor-^r-^csih-^CMcridcMcOT-coddr^-cboc)
ininininiDininTrcMCMTr'-cNcDinin'^M-cMCMT-
incoincDin'vcoT-'" q-cocM^'r-'-"m'3-ininro
0)
a
> > > > >
2 2 2 2 2
Q- CL D- D-
X
CO
CD
CO
CD
^
C C C C
CD
CO
CD
CO
c c c c c
CL D. D. Q.
> > > >
2 2 2 2
'=222222in
CD
CD
Q. 0.
^^^ ^ ^ ^ > >
CD
=
CD
CD
CD
CD
ni
(D
"c "c ~C
CD
c
CD
00
co(X)coin'^r-.a)h^r^a>0)r--coa)f^tDcoou^oococoT-T-rO'»--^fooooior^
r^J(N^^iGOO^cocococococooooococOfO<TTr(n(Nror^cocr)aocococooooooococo^^
I
<
I
^'-^OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
OOCDOOOOOOOOOOOOOCDOO
0 0 0 0 0 0 0 0 0
Q.
O
C
CD
CD
CD
CO
CO
CO
CO
CO
CDCDCOCDCDCDCOCO
LLLLLLU-LLLLLLU.
CD
LL LL U. U- U- LL
00000 0000
0000
CM CM CM CM CM
<t in CD
0000
o o
in in
in CD CD CD
0000000 000000
o o o o o
CNJ
in
CM CM CM CM CM CM in CD CO
CD CD
CD
CO
CD
CD
(D
CO
CD
000000
o o o o o
CO
in
in CD CO CD
Q. It
C
= „
CD
,9
o
ra
o -a m
u E
CD
CO
CD
CD
CD
^^^^^ctrcrcfcfcrcrcfcccicSccicc
^cc
COCDCDCDCDCDCOCOCDCDCD^^
cc'c'ccc'c'c'cc'c'c'c'c'c
CD
<D
OOOOOOOOCDCDCDCDCOCOOOOOOOOOOOOOOOOOOOOCOCDCOCDCDCDCDCD
ininininLnLninincocococococoinininininininininintninininininininincococococococoro
—
NISTNCSTAR
1-3D,
WTC
Investigation
191
iiiiinmunuM
lllllllllllllllllljiiijiiiiiiiiiiiii
o o o o o
0(OCMoooTcn(_jiocooaDQOoniDiD'-cs]oooinoc\ioocoOLn^n(0(>Jor^jj,
45.4
38.5
o in
^5
38.3
47.0
37.3
27.8
40.9
a>
(o
?8
20.9
^ ^ D>
ra§ m
c
c c
ooooooooooooooooooooooo
_i_l_l|-l-l-(--J_J|--J-lH-l-l-l-i-l-i(-l-l-l-l-l-l-(-(--l-l-l|-l-l-l-H
§SS§§S§goS°oS
CM
r\j
rvj
CM
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
Q.
E
TO
CD
N-r-.'*'<Too-<-'^rMT-T-cNj^coLr)c\jCNJc\Jc£ir^uo
<<<<<<<<<<<<<<<<<<<<
l-l-l-l-l-l-l-l-l-l-l-l-l-l-l-l-l-l-l-l-
CNCMCMCNCNCMCN(NCMCMOJCMCNI(MC\JCNC\JC\iCMrvJ
OOOOOOOOOOOOOOOOOOOO
0 0) 0
3 3 D 3
C C C C
0
C
Q)
0
C
0}
C
SS55<150000^0^0000Efl300
000033333030333303D3
= === = ===
« = = = =
:=:=-^:z:n3ro03ro03:^ro^:;n303n303 -^030303
C/)
t/3
C/5
c/D
c/5
r^a5ncoocoa)tDOT-T-cDLncr)T-o-<^cNja5oo
a
E
0)
CX3CNJCCa5C\jrOOO''--^OOOI^OOCOCOT--r--;tCNOO
OLnoocNja3ooooir)oo-<^T-';i-cr)r^a50Lnoooon
OOOOOOOO-'-OT-OOOO'^OrvJ-r-;-^
oooooooooooooooooooo
(0
3
5 ooog^g
O
»-
5g^
a,
5
CO
00
m
CD CO
CM
CO
3
10
«
LnOCOO-^CD-<-CDOLnCDOOCMO-<-OOCOCDir>
cT>-<d-c7)CNj->^LnT-ooa5cnoLr)'<-cooococ75cy)C35
unco-<;f'<3-CNjcor^cDt^O'^'*coc»i^ootDm'^cj)
cocDoocD(3^c3or^t^i^r^jcocvJOooh--c3^CNjooc3jr~-
OOOCDCDCDCZiciio-r-^'^-r^-^-"— r^CZi-^CDOCNj
(0
o
LDLD'd-Lnirj'^'^-^cDi^cDmcjicoLnuo'*-^-^
a
o
u
o
CD
OOOOOOOOOOOOOOOOOO-"—
Ljj
UJUJUJLJJljjlJJujljjljjuJljjLlJIjjmUJUJL^
CO
o
CN
D0
OC\IOLOOO'^C75C\ICOCJ5CD(N'^ir)C\l^--h--OtD
C
o
c
w
o
'c
OCOv-|^OrMLncr>CNOOC30-<-CNC75COCMCnOOO
-,-
-I-
'T-^CNi-^CNic»(NCsiCNjcNicTjT--^C35-5-cb'^t^
'co
c
©
CJ
CO
c\icot-^cououocoou:)C3:)r-aDcqoqunT-coT-->;j-^
0'r^Trcbcbcj)cz>c6cor^cj)-r-^c\icO'^o6oc5-<-T-^
k
CO
Q.
T-OCT)C35OOCJ5O^-<^C3^OCT>LnU0OCDOh~
U0CDCDCOCOCO'3-->:J-'^T-T-CNJCslC\jT-T-T-C\jrM"-'
as
D)
eri
nued
3
*>
ine
00'<*'<d-CD'<^00C0C0CT>OC0CD-^r--C0rrCM-^T-C0
r^oooor-^oo^-oocor-oooocooor^coc30ooc»c30oo
re
< —
a>
a
C7)
c 0 c
o
0
o o _c
T3 •o c
0
OOOOOOOOOOOOOOOOOOOCD
OOCDOCDCDOOOOOCDCZioczjcDcicDCDCD
CNJCMCNJCNJCvJCNJCNJCNJCNJCMCNJCNJCNJCNJCMCMCNCvJCMCM
i
C/3
E
CO
^
JZ
O)
'£
oooooooooooooooooooo
III
E
I
I
1
I
I
^5i5i5i5i2i5i5i5i2i5i5i5 — — — — — —
I
0
i5
03
o0
oooooooooooooooooooo
ooooooooooooooLntnunLnino
<d--^^muniDLnLnLOcDCDcoc£)cDCDtDcDCDCDr^
©
Str
c
0
im
o O
0
ULLJ_LLLLLLU-LJ-U_Ll_U_U_lJ_IJ_U_Ll_U_U_LJ_LJ_LL
d
'cD
wa
CSJCNCNICM(>JCMC\J(NCNJCNlC\irMCMCslCMCMCNJCNJCMC\J
(/)"
'o
CO
Cl "E
(O
ich
as
c
0
0
i_
o c 0
0
x:
re
LU!
03
sp
>-
tra
(/)
o
_
_ c
0 r-
^'
<1>
<u
>
Ct3
cn
Q
03
03
*
wh
erin
CC
ne
03CDCDCDCT3CDCD03CDCDCDCt3CD_ro_ro_CD_ro
At
=
O 3 o
0
E c
0
CD
n
'cD
1
to
D
oooooooooooooooooooo
lOunu^ioinunuoioioirjunLnLOLnLnLnLnLOLnLn
CO
NISTNCSTAR
>
1-3D,
WTC
Investigation
E
CO
0
O
J"
0
193
Appendix A
A.5
Figures
SUPPLEMENTAL FIGURES FOR ROOM-TEMPERATURE PROPERTIES
A-1 through A-17 summarize
the room-temperature stress-strain behavior of specimens taken
from the perimeter columns. Figures A-1 8 through A-26 sununarize the room-temperature
behavior of specimens taken from the core wide-flange and box columns. Figure
behavior of the truss
stress-strain
A-27 summarizes
the
steels.
X
S9-C1 T-S1
WTC
—1—
I-
1
Col 133 97-100
1
1
1
1
1-- 1—
1—r-1
1
—
r
TT'-i
—r—
Specified = 36 ksi
Ifj
1/
'1/
s9-c1t-s1-l1
s9-c1t-s1-12
s9-c1t-s1-l3
s9-c1t-s1-t1
s9-c1t-s1-t2
'
1
0
005
0
.
.
.
.
010
1
.
.
.
0 015
Engineering Strain
J
0.00
0.05
0.10
I
I
I
0.15
I
I
I
I
0.20
0.25
Engineering Strain
15:01:41
Sunday September 12.2004
Figure A-1. Stress-strain curves for Fy = 36 ksi S-9 spandrel.
194
NISTNCSTAR
1-3D.
WTC
Investigation
Data Tables and Supplemental Figures
Figure A-2. Stress-strain curves for Fy = 45 ksi
ASCE3.
Figure A-3. Stress-strain curves for Fy = 46 ksi S-14 spandrel.
NISTNCSTAR
1-3D.
WTC
Investigation
195
————— ——
-|
I
I
I
I
I
—— ——— —
I
I
I
I
I
I
"T
80
M26-C1T-RF
WTC
w
1
Col 131 93
60
-4—
C/D
O)
i
40
<D
<D
C
c
LU
m26-c1b1-rf-11
m26-c1b1-rf-l2
20
iTi26-c1bi-rf-!1
m26-c1b1-rf-t2
0 000
0 005
0 010
015
0
Engineering Strain
J
0.00
0.05
±
l_
0.10
_1_
J
0.15
I
l_
I
0.20
0.25
Engineering Strain
09:34:55
-
Friday October 29,
2004
Figure A-4. Stress-strain curves for Fy = 50 ksi M-26.
— —— —
T
I
I
— — ——i— — — — — — —
-1
I
0 000
I
I
1
I
I
I
I
0 015
0 005
I
I
0 020
Engineering Strain
0
-I
0.00
I
I
I
0.05
I
1_
I
0.10
0.15
I
I
I
0.20
I
I
i
I
I
0.25
1
0.30
Engineering Strain
11:51:19
-
Sunday September
12,
2004
Figure A-5. Stress-strain curves Fy = 50 ksi CC-C3T-IW.
NISTNCSTAR
1-3D,
WTC Investigation
r
r
Data Tables and Supplemental Figures
T
——
———
T
1
1
I
I
n
I
I
I
80
N9-C1T1-RF
WTC
w
Col 155 102
1
60
03
(D
i_
'
.
'
1
1
1
1
1
1
CO
F, Specified
i
40
n9-c1t1-rf-l2
-
1
T
c
1
.
1
T
c
;
LLI
20
ksi
i
n9-cit1-rf-t1:
CD
= 55
/
:
i
/
i
•f
1
,
,
,
.
1
0
,
005
.
.
,
0 010
1
.
,
.
,
0 015
Engineering Strain
0
J
0.00
0.05
0.10
I
I
I
0.15
I
I
I
I
J
I
I
0.20
[_
0.25
Engineering Strain
15:05:47
-
Sunday September
12.
2004
Figure A-6. Stress-strain curves for Fy = 55 ksi N-9.
Figure A-7. Stress-strain curves for Fy = 55 ksi N-9 inner web.
NISTNCSTAR
1-3D.
WTC
Investigation
197
Figure A-8. Stress-strain curves for Fy = 60 ksi N-8.
———
I
n
I
-i
I
P
i
I
I
I
F
n8-c1b1-a-iw-l1
80
N8-C1B1-IW
WTC
CO
O)
i
1
Col 143 98
60
Specified = 60
l<si
40
CD
CD
g
^
20
0 000
0
_1_
005
0
0 015
010
0
020
Engineering Strain
J
0.05
0.00
I
1
L
0.10
I
I
I
I
I
I
0.15
I
I
I
I
I
I
L.
0.20
0.25
Engineering Strain
15:33:43
-
Sunday September
12.
2004
Figure A-9. Stress-strain curves for Fy - 60 ksi N-8 inner web.
NISTNCSTAR
1-3D,
WTC
Investigation
r
Data Tables and Supplemental Figures
100
—— —
I
T
I
N99-C3T1-RF
WTC
1
Col 147 102
n99-c3t1-rf-l1
n99-c3t1-rf-l2
n99-c3t1-rf-t1
n99-c3t1
-rf-t2
0 000
0 005
0010
0
015
0
020
0 025
Engineering Strain
0.00
0.05
0.10
0.15
0.20
0.25
Engineering Strain
15:09:41
-
Sunday September
12.
2004
Figure A-10. Stress-strain curves for Fy = 65 ksi N-99.
Figure A-11. Stress-strain curves for Fy = 70 ksi C-46.
NISTNCSTAR
1-3D,
WTC
Investigation
199
Appendix A
100
— — — — — — — — — — —i—
I
I
I
I
I
I
I
I
I
i
-i
80
I
I
I
I
s1-c2m-fl-2-l1
s1-c2m-fl-2-l2
s1-c2m-fl-2-t1
s1-c2m-fl-2-t2
CO
C/D
0)
60
1
1
1
1
r
1
1
1
1
1
r
1
1
I
1
1
r
1
1
CO
S1-C2M-LF
Specified = 70 ksi
c
WTC
1
Col 433 80
40
CD
0)
c
-
"O)
c
Il
/'/
UJ
:///
20
//
y
'
.
,
.
.
1
0015
0 010
0 005
Engineering Strain
0
_J
I
L _1
I
0.00
I
I
I
I
L.
I
I
0.05
_l
I
I
I
I
L.
I
I
0.15
0.10
0.20
0.25
Engineering Strain
09:47:15
-
Friday October 29, 2004
Figure A-12. Stress-strain curves for Fy = 70 ksi S-1.
100
n
I
I
I
I
80
r
!
1
s14-c3m-fr-1
s14-c3iTi-fr-1-l2
s14-c3m-tr-i-t1
s14-c3m-fr-1-t2
in
w
60
.
.
.
.
.
,
CO
-
J-'
Specified = 70 ksi
-
0)
<D
1
40
S14-C3M-RF
WTC
c
D)
C
2 Col 218 92
LU
20
.
.
0 005
0.000
1
.
,
.
,
0 010
1
.
.
,
.
0 020
0 015
Engineering Strain
I
0.00
I
\
I
0.05
I
\
\
I
I
0.10
1
I
\
I
L
0.15
0.20
0.25
Engineering Strain
09:58:34
Friday October 29,
2004
Figure A-13. Stress-strain curves for Fy - 70 ksi S-1 4.
200
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
0.00
0.05
0.10
0.15
0.20
0.25
Engineering Strain
1536:42 -Sunday September
12,
2004
Figure A-14. Stress-strain curves for Fy = 70 ksi S-14 inner web.
Figure A-15. Stress-strain curves for Fy = 75 ksi C-22.
NISTNCSTAR
1-3D.
WTC
Investigation
201
Appendix A
Figure A-16. Stress-strain curves for Fy = 80 ksi C-25.
— — — — — ——— — — ——— —— — — —— — — — —
I
I
I
I
I
T
I
I
I
0 005
0,000
I
I
I
I
I
0 015
0.010
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0.020
Engineering Strain
0
I
\
I
0.00
I
I
I
0.05
I
I
I
I
I
0.10
I
I
I
I
I
I
0.15
I
0.20
I
0.25
Engineering Strain
10:14:18
-
Friday October 29,
2004
Figure A-17. Stress-strain curves for Fy = 100 ksi C-10.
202
NISTNCSTAR
1-3D,
WTC Investigation
Data Tables and Supplemental Figures
Figure A-18. Stress-strain curves for core wide-flange column C-30.
NISTNCSTAR
1-3D,
WTC
Investigation
203
Appendix A
Figure A-20. Stress-strain curves for core wide-flange column C-71.
Figure A-21. Stress-strain curves for core wide-flange column C-155.
204
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
80
60
HH 12WF92
WTC
Col 605 98-101
1
w
o
CO
40
a3
o
c
D)
C
hh-fl-1-2-l1 flange
20
LU
hh-fl-1-2-12 flange
hh-fl-1-16 flange
hh-fl-i-2-tl flanae
hh-fl-1-2-t2 flange
0015
0 010
0 005
Engineering Strain
J
0.00
L
I
0.04
0.02
\
<
<
^
0.06
L.
\
0.08
0.10
Engineering Strain
10:57:48
-
Fnday October 29 2004
Figure A-22. Stress-strain curves for core wide-flange column HH.
80
-|
c26-f2-1-1
1
r
web
_i_
0 000
_i_
0 005
0 015
0,010
0 020
Engineering Strain
_l
0.00
0.02
1
I
I
0.04
I
I
I
I
1
0.06
1
I
I
I
0.08
I
L
0.10
Engineering Strain
Figure A-23. Stress-strain curves for core wide-flange
NISTNCSTAR
1-3D.
WTC
Investigation
beam
C-26.
205
Appendix A
70
iange)
B6152-1-F1-1-15 (flange)
60
B6152-1 box column
1 Col 803 15-18
50
WTC
CO
o
^
y
40
c
0)
CD
———
30
-I-
Specified = 36 ksi
/
c
"CD
c 20 r
LU
J
1
V
10
1
0.000
,
.
,
,
0010
0.005
1
,
.
.
,
0 015
Engineering Strain
J
J
L_
1
0.00
I
L-
\
1
1
0.04
0.02
\
I
I
I
0.06
I
0.08
0.10
Engineering Strain
11:34:31
-
Friday October 29, 2004
Figure A-24. Stress-strain curves for core box column B6152-1.
70
B6152-2-F1-2-I1
B6152-2-F1-2-!2
B6152-2-F1-2-t1
B6152-2-F1-2-t2
60
(0
^
B6152-2 Box column
1 Col 504 33-36
50
WTC
CO
C/3
CD
CD
C
'g^
20
LU
10
0 005
0 015
0 010
Engineering Strain
J
I
0.00
J
L.
0.02
I
L
I
0.04
I
1
0.06
I
I
I
0.08
0.10
Engineering Strain
1 1
33:05
-
Friday October 29.
2004
Figure A-25. Stress-strain curves for core box column B6152-2.
206
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
Figure A-26. Stress-strain curves for core box column C-88c.
0.00
0.05
0.10
0.20
0.15
0.25
Engineering Strain
11:39:13
•
Friday October 29 2004
Figure A-27. Stress-strain curves for
NISTNCSTAR
1-3D,
WTC
Investigation
all
truss angles.
207
Appendix A
SUPPLEMENTAL FIGURES FOR HIGH STRAIN-RATE PROPERTIES
A.5
A-28 through A-37 summarize the stress-strain curves as a function of strain rate for perimeter
column steels. Figures A-44 through A-52 summarize the curves for core-coluinn steels. Figures A-53
through A-68 summarize the strain rate sensitivities for perimeter column steels. Figures A-69
through A-77 suminarize the strain rate sensitivities for core column steels. Figure A-78 through A-80
summarize the dependence of total elongation, El„ on strain rate. Figures A-81 through A- 103 show the
Figures
results
208
of high
strain-rate
Kolsky
tests.
NISTNCSTAR
1-3D,
WTC Investigation
Data Tables and Supplemental Figures
M26-RF 50
ksi (Longitudinal)
120
-800
-600
(A
«
80-
10
0)
re
Q.
>*-•
CO
-400
c
0)
0)
c
5,
Fy strain Rate,
40-
6.06
c
E
s"^
-5
65
HI
-200
100
260
417
\
0.1
\
I
0.2
0.3
Engineering Strain,
0.4
in/in
Figure A-28. Stress-strain curves for Fy = 50 ksi perimeter column M26-C1B1-RF
longitudinal.
M26-RF 50
ksi
(Transverse)
120
-800
-600
V)
V)
80-
re
Q.
(0
O)
-400
c
'!!
0)
<U
c
Fy Strain Rate, s
40-
6.06
c
99
liJ
E
-5
-200
255
316
360
0
1
0
\
0.1
\
0.2
Engineering Strain,
0
\
0.3
0.4
in/in
Figure A-29. Stress-strain curves for Fy = 50 ksi perimeter column M26-C1B1-RF
transverse.
NIST NCSTAR
1-3D,
WTC
Investigation
209
Appendix A
N8-RF 60
ksi (Longitudinal)
800
-600
03
Q.
-400
0)
0)
Fy Strain Rate, s
•I 40
H
1.07 E -4
c
90
-200
257
277
386
n
0.1
1
\
0.2
0.3
Engineering Strain,
T"
0.4
0.5
in/in
Figure A-30. Stress-strain curves for Fy = 60 ksi perimeter column N8-C1B1-RF
longitudinal.
N8-RF 60
ksi (Transverse)
800
100-
-600
(A
750)
cn
C
=
O
c
-400
50-
"5)
111
F„ Strain Rate,
25
s
-200
0.005
54
297
0.1
0.2
Engineering Strain,
0.3
0.4
in/in
Figure A-31. Stress-strain curves for Fy = 60 ksi perimeter column N8-C1B1-CRFtransverse.
210
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
N99-RF 65
ksi (Longitudinal)
150
-1000
-800
V)
w 100
V)
H
<3)
600
(0
TO
Q.
-400
0)
c
ui
-200
T
0
0.2
0.1
0.4
0.3
Engineering Strain,
in/in
Figure A-32. Stress-strain curves for Fy = 65 ksi perimeter column
N99-C3M1-RF
longitudinal.
N99-RF 65
ksi (Transverse)
120
-800
CA
[-600
BO0)
Q.
D)
_C
-400
a>
E
o
c
40
H
Fy Strain Rate,
s-^
6.06 E -5
LU
79
-200
270
515
0
0.1
0.2
Engineering Strain,
0.3
0.4
in/in
Figure A-33. Stress-strain curves for Fy = 65 ksi perimeter column N99-C3M1-RF
transverse.
NISTNCSTAR
1-3D,
WTC Investigation
211
Appendix A
C22-FL 75
ksi (Longitudinal)
150
-1000
-800
(A
(ft
100-
U)
O
-600
OT
S)
™c
0)
-400
c
50^
cn
c
LU
-200
0.1
0.2
Engineering Strain,
in/in
Figure A-34. Stress-strain curves for Fy = 75 ksi perimeter column C22-C2B-FL
longitudinal.
C22-FL 75
ksi (Transverse)
150
- 1000
-800
(A
to
(A
0)
100
[-600
ni
O)
Q.
-400
Fy strain Rate,
c
8.33
HI
121
E
s""
-5
-200
276
358
0
\
\
0.1
0.2
Engineering Strain,
0.3
in/in
Figure A-35. Stress-strain curves for Fy = 75 ksi perimeter column C22-C2B-FL
transverse.
212
NISTNCSTAR
1-3D,
WTC Investigation
*
Data Tables and Supplemental Figures
C14-SP 75
ksi (Longitudinal)
150
- 1000
-800
w 100-1
v)
k.
-600
(O
D)
(0
Q.
C
'Z!
0)
o
400
i 50H
C
LU
-200
0-]0
T
\
0.1
0.2
Engineering Strain,
0.3
in/in
Figure A-36. Stress-strain curves for Fy - 75 ksi perimeter column spandrel C14-SP
longitudinal.
C14-SP 75
ksi (Transverse)
1000
120-
800
(/)
(/T
(/)
0)
-600
*—
D)
C
'l.
400
0)
_c
LU
-200
0
0.1
0.2
Engineering Strain,
in/in
Figure A-37. Stress-strain curves for Fy = 75 ksi perimeter column spandrel C14-SP
transverse.
NISTNCSTAR
1-3D,
WTC
Investigation
213
Appendix A
M10B-RF 100
ksi (Longitudinal)
1200
-1000
.
120-
-800
(A
0)
to
-600
.E
0)
0)
c
-400
"5)
c
Hi
-200
0.05
0.1
0.15
Engineering Strain,
0.2
in/in
Figure A-38. Stress-strain curves for Fy = 100 ksi perimeter column
M10B-C3B1-RF
longitudinal.
M10B-RF 100
ksi (Transverse)
1000
-800
-600
-400
-200
0
0.1
0.2
Engineering Strain,
in/in
Figure A-39. Stress-strain curves for Fy = 100 ksi perimeter column M10B-C3B1-RF
transverse.
214
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
C10-FL 100
ksi (Longitudinal)
1200
160-
-900
(A
120
a.
CO
-600
.E
80
H
0)
0)
Fy Strain Rate, s
c
"o
UJ
E-4
1.1
c
-300
70
40-
^
241
325
433
I
\
0.05
0.1
I
I
0.15
0.2
Engineering Strain,
0.25
in/in
Figure A-40. Stress-strain curves for Fy = 100 ksi perimeter column C10-C1M1-FL
longitudinal.
C10-FL 100
ksi
(Transverse)
1200
160-
-900
120V)
V)
c
E
-600
80-
0)
c
UJ
300
40-
0
0.1
0.2
Engineering Strain,
in/In
Figure A-41. Stress-strain curves for Fy = 100 ksi perimeter column C10-C1M1-FL
transverse.
NISTNCSTAR
1-3D.
WTC
Investigation
215
Appendix A
C10-IW 100
ksi (Longitudinal)
150
-1000
-800
-600
-400
-200
T
0.05
0.1
0.15
Engineering Strain,
0.2
0.25
in/in
Figure A-42. Stress-strain curves for Fy = 100 ksi perimeter column C10-C1M1-IW
longitudinal.
C10-IW 100
ksi
:
(Transverse)
r
1200
160-
-900
:
120-
V)
0)
k.
-600
U)
c
o
o
c
'Ui
c
-300
HI
1
0
0.05
\
0.1
r
1
0.15
Engineering Strain,
0.2
0.25
in/in
Figure A-43. Stress-strain curves for Fy = 100 ksi perimeter column C10-C1M1-IW
transverse.
216
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
C65-FL 36
ksi (Longitudinal)
80-
60-
-400
in
o>
(O
Q.
c
^400)
0)
c
Fy Strain Rate, s
"o)
c
m
200
^
8.75 E -5
20-
81
238
^307
~1
1
0.2
0.1
T"
\
0.4
0.3
Engineering Strain,
0.5
in/in
Figure A-44. Stress-strain curves for Fy = 36 ksi core wide-flange C65-FL longitudinal.
C65-FL 36
ksi (Transverse)
800
100
-600
V)
(/)
75-
CO
c
-400
50-
Q)
C
Ui
c
LU
-200
25-
1
0
0.1
I
0.2
Engineering Strain,
r
0.3
in/in
Figure A-45. Stress-strain curves for Fy = 36 ksi core wide-flange C65-FL transverse.
NISTNCSTAR
1-3D,
WTC Investigation
217
Appendix A
C65-WEB 36
ksi (Longitudinal)
90
-600
-400
ns
CL
-200
~i
0.1
1
0.2
r
\
0.4
0.3
Engineering Strain,
0.5
in/in
Figure A-46. Stress-strain curves for Fy = 36 ksi core wide-flange
C65-WEB
longitudinal.
V
C65
WEB 36 ksi
(Transverse)
120
-800
-600
(A
80<fl
(A
Q)
1-
V)
ra
-400
0)
0)
c
'u>
40-
c
HI
-200
0.1
0.2
Engineering Strain,
in/in
Figure A-47. Stress-strain curves for Fy - 36 ksi core wide-flange
218
C65-WEB
NISTNCSTAR
1-3D,
transverse.
WTC Investigation
Data Tables and Supplemental Figures
C80 36
ksi (Longitudinal)
800
100-600
80
w
o
J
(/)
60-
-400
c
0)
c
40-
'O)
c
-200
LU
20-
T
0
0.2
0.4
Engineering Strain,
0.6
in/in
Figure A-48. Stress-strain curves for Fy = 36 ksi core wide-flange C80-A longitudinal.
HH1-WEB 42
ksi (Longitudinal)
800
6.3
E
-4
68
233
0
0.05
0.1
0.15
Engineering Strain,
0.2
0.25
in/in
Figure A-49. Stress-strain curves for Fy = 42 ksi core wide-flange
NIST NCSTAR
1-3D.
WTC
Investigation
HH1-WEB
longitudinal.
219
Appendix A
HH1-WEB 42
ksi (Transverse)
800
-600
80(A
S
(0
O)
-400
c
'!Z
o
0)
c 40-
'5)
c
-200
UJ
0.1
1
r
0.2
0.3
Engineering Strain,
in/in
Figure A-50. Stress-strain curves for Fy = 42 ksi core wide-flange
B6152-FL 36
HH1-WEB
transverse.
ksi (Longitudinal)
120
-800
-600
re
Q.
400
F„ Strain Rate,
s-^
-200
"~l
0.1
T
T
T
0.2
0.3
0.4
Engineering Strain,
0.5
in/in
Figure A-51. Stress-strain curves for Fy = 36 ksi core box column B6152-FL longitudinal.
220
NISTNCSTAR
1-3D.
WTC Investigation
—
.
Data Tables and Supplemental Figures
B6152-FL 36
ksi (Transverse)
120
-800
-600
-400
-200
1
r
0.1
0.2
0.3
Engineering Strain,
0.4
in/in
Figure A-52. Stress-strain curves for Fy = 36 ksi core box column B6152-FL transverse.
2.2
-
2.1
—
160
1
1
I
1
1
o
"
1
1
1
1
1
1
1
1
1
m=0.01 21 +0.0033 F Longitudinal
nn=0.01 1 8±0.0023 TS Longitudinal
140
120
V)
c
100 i=
2.0
CD
c
W
0}
o
S
^
1.9
]
1
1
-4
-
-
'
Wednesday Febmary
1
1
-2
1
1
-
80
—o
Perimeter column
M26-C1B1 F =62,
1.8
19:43:12
—
o
^
O
ksi
60
1
0
2
log.io strain rate, 1/s
09,
2005
Figure A-53. Strain rate sensitivity for Fy = 50 ksi perimeter column M26-C1-RF
longitudinal.
NISTNCSTAR
1-3D,
WTC
Investigation
221
Appendix A
160
140
120
CO
100
B
c
0
-2
log-,0
19:43:13
-
Wednesday February
09.
strain rate, 1/s
2005
Figure A-54. Strain rate sensitivity for Fy = 50 ksi perimeter column M26-C1-RF
transverse.
160
2.2
140
- 120
- 100
i:
C
CD
CO
-2
0
log-io strain rate,
19:43:13
-
Wednesday February
09.
1/s
2005
Figure A-55. Strain rate sensitivity for Fy = 60 ksi perimeter column N8-C1-RF
longitudinal.
222
NISTNCSTAR
1-3D.
WTC
Investigation
Data Tables and Supplemental Figures
Figure A-56. Strain rate sensitivity for Fy = 60 ksi perimeter column N8-C1-RF
transverse.
Figure A-57. Strain rate sensitivity for Fy = 65 ksi perimeter column N99-C3-RF
longitudinal.
NISTNCSTAR
1-3D.
WTC
Investigation
223
Appendix A
Figure A-58. Strain rate sensitivity for Fy = 65 ksi perimeter column N99-C3-RF
transverse.
2.2
^
1
1
1
1
1
1
O
r
1
1
160
1
1
1
1
_
m=0.0067±0.0021 F„ Longitudinal
m=0.01 38x0.0003 TS Longitudina!
o
140
2.1
120
V)
§
100 -h
2.0
—
CO
'
o
c
O
(D
o
'
C/D
^
80
1.9
Perimeter column
1.8
C^2-C2M Fy=83
.
1
1
1
1
1
-2
-4
1
1
Ifsi
60
L.
0
2
log^o strain rate, 1/s
19:43:14
Wednesday February
09,
2005
Figure A-59. Strain rate sensitivity for Fy = 75 ksi perimeter column C22-C2M-FL
longitudinal
224
NIST NCSTAR
1-3D.
WTC
Investigation
—
—
Data Tables and Supplemental Figures
2.2
r
1
1
1
1
1
1
1
1
1
1
O
160
1
1
1
^
m=0.0061±0.0016 F^, Transve.rse
m=0.0145±0.0038 TS Transverse
140
2.1
120
^
-
——
-
"
- 100
'
o
cn
c
-
0)
"
o
TO
G-
-
^
80
1.9
Perimeter column
1.8
C22-C2M F =83
-
1
,
,
,
1
,
,
60
-2
-4
ksi
,
2
0
log^Q strain rate, 1/s
19:43:14
•
Wednesday February
09,
2005
Figure A-60. Strain rate sensitivity for Fy = 75 ksi perimeter column C22-C2M-FL
transverse.
2.2
-
1
1
1
1
1
1
1
1
1
1
1
1
160
1
1
0
F^,
Longitudinal
m=0.0109±0.001 5 TS Longitudinal
140
^
2.1
—-o
(5
0" — 120
CO
%
c
0 - 100
0
2.0
\
i=
CD
c
W
CD
o
(J)
d)
^
80
1.9
Perimeter column
1.8
CH-SP
>
1
1
1
1
1
-2
19:43:16
•
Wednesday February
1
1
F,=92
ks)
,
L_
60
0
loQio Strain rate, 1/s
09,
2005
Figure A-61. Strain rate sensitivity for Fy = 75 ksi perimeter spandrel C14-SPlongitudinal.
NISTNCSTAR
1-3D,
WTC
Investigation
225
Appendix A
T
2.2
O
2.1
160
r
1
m=0.0083±0,0021 F,, Transverse
m=0.0098±0.00l 6 TS Transverse
140
120
CO
CO
100 i=
O)
c
cn
CD
o
^
80
1.9
Perimeter column
1.8
C14-SP F =92
J
I
60
-2
Wednesday February
19:43:16
09.
kS|i
L.
0
log 10 strain rate, 1/s
2005
Figure A-62. Strain rate sensitivity for Fy = 75 ksi perimeter spandrel C14-SP transverse.
1
2.2
O
2.1
1
1
r
1
r
T
1
160
r
n-i=0.0053±0.0014 F„ Longitudinal
m=0.0085±0,001 3 Tfe Lonaitudmal
140
120
- 100
CD
0)
CO
o
80
1.9
Perimeter column
M10B-C3B1 F =1,08
1.8
J
I
-4
L.
_l
-2
I
ksi
60
L
0
log^Q strain rate, 1/s
19 43:14
•
Wednesday Febmary
09,
2005
Figure A-63. Strain rate sensitivity for Fy = 100 ksi perimeter column
M10B-C3B-RF
longitudinal.
226
NISTNCSTAR
1-3D.
WTC
Investigation
Data Tables and Supplemental Figures
160
2.2
m=0.0039±0.0038 F^. Transverse
m=0.0083±0.0010 TS Transverse
140
2.1
120
100 i=
C
4—*
W
CD
o
TO
^
80
1.9
Perimeter column
M,10B-C3B1 F =1,08
I
-4
19:43:15
-
-2
Wednesday February
I
I
LI
I
ksi
I
60
0
log.io strain rate, 1/s
09,
2005
Figure A-64. Strain rate sensitivity for Fy = 100 ksi perimeter column M10B-C3B-RFtransverse.
160
-|
2.2
1
O
I
m=0,0044±0.0014 F Longitudinal
m=0.0080±0.0026 TS Longitudinal
140
2.1
- 120
CO
0)
cn
c
100 -B
2.0
CD
CD
W
CD
C
o
TO
^
80
1.9
Perimeter column
=110,
C10-C1M F
-J
I
-2
19:43:15
•
Wednesday February
ksi
60
L.
0
loQio Strain rate, 1/s
09.
2005
Figure A-65. Strain rate sensitivity for Fy = 100 ksi perimeter column C10-C1M1-FL
longitudinal.
NISTNCSTAR
1-3D.
WTC
Investigation
227
Appendix A
2.2
F
—— ——
n
'
1
'
'
O
I
160
F
F„ Transverse
ife
Transverse
140
2.1
o
o
-
120
S
100
2.0
C
0
C/D
^
- 80
1.9
Perimeter column
1.8
C10-C1M-IW F =110
_J
I
ksi
60
L
-2
0
log^o strain rate, 1/s
19:43:16
-
Wednesday February
09.
2005
Figure A-66. Strain rate sensitivity for Fy = 100 ksi perimeter column C10-C1M1-FL
transverse.
2.2
F^—
O
1
1
1
r
m=G.0077±0.0033
F.,
1
160
r
Longitudinal
=0.0113±0,0006 TS Longitudii
140
2.1
- 120
CO
g
100
2.0
•I—
C
cn
CD
CO
^
- 80
1.9
Perimeter column
1.8
C10-C1M-IW F =110
_i
I
[_
_l
-2
I
ksi
60
L
0
log^o strain rate, 1/s
19:43:16
-
Wednesday Febmary
09.
2005
Figure A-67. Strain rate sensitivity for Fy = 100 ksi perimeter column C10-C1M1-IWlongitudinal.
228
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
T
2.2
r
1
-1
O
F,
o
TS Transverse
1
160
r
Transverse
140
o
o
2.1
120
(0
i
2.0
100 i=
c
o
O
80
1.9
Perimeter column
C10-C1M-IW F =110
-I
L.
I
I
-2
19 43 16
-
Wednesday February
'
I
'
I
I
ksi
i__
60
0
log^o strain rate, 1/s
09.
2005
Figure A-68. Strain rate sensitivity for Fy = 100 ksi perimeter column C10-C1M1-IW
transverse.
1
1
I
[
1
1
1
1
1
1
1
1-
1
1
2.0
-
o
m=0.0290±0.0042
m=0.01 3010.001 4
D
F.,
100
Longitudinal
Tfe Longitudinal
80
C/3
o
—1
CO
o
CD
C
60
c
0
o
§^1.6 —
'"Q^
40
O
§
Core column
C^SJIange f^=3^
1.4
1
1
1
-4
19:48:00
-
^
Wednesday February
1
1
-2
1
1
ksi
1
0
2
log.io strain rate, 1/s
09.
2005
Figure A-69. Strain rate sensitivity for Fy = 36 ksi core wide-flange C65-FL longitudinal.
NISTNCSTAR
1-3D.
WJC
Investigation
229
Appendix A
-|
2.0
O
r
1
n
r
I
100
F Transverse
TS Transverse
80
CO
1.8
V)
60
^
c
0
c
CO
CD
o
O
40
1.6
Core column
C^SJIange
1.4
_l
I
-4
-2
0
log.iQ strain rate,
19:48:01
-
Wednesday February
09,
Fy=3:^ ksi
L.
1/s
2005
Figure A-70. Strain rate sensitivity for Fy - 36 ksi core wide-flange C65-FL transverse.
1
2.0
O
1
r
100
m=0.0336±0.0030 F Longitudinal
m=0,0207±0.0010 TS Longiludinai
80
CO
1.8
60
cn
c
CD
c
cn
CD
40
Core column
1.4
J
I
C^5_web Fy=31
L
-2
log.io strain rate,
19 47;59
-
Wednesday February
09,
Ifsi
0
1/s
2005
Figure A-71. Strain rate sensitivity for Fy = 36 ksi core wide-flange
230
C65-WEB
NISTNCSTAR
1-3D.
longitudinal
WTC Investigation
Data Tables and Supplemental Figures
1
1
1
1
[
2.0
~
1
[
O
F Transverse
o
TS Transverse
T"
100
°
°o
:
80
oo
'
-
o
1.8
CO
60
^
c
C
(D
§^1.6
40
Core column
C^5_web Fy=31
1.4
1
1
1
1
1
1
1
Ifsi
1
-2
0
log.,Q strain rate, 1/s
19:47:59
•
Wednesday Febnjary
09.
2005
Figure A-72. Strain rate sensitivity for Fy = 36 ksi core wide-flange
-1
2.0
O
1
transverse.
r
100
m=0.0317±0.0027 F Longituciinal
m=0.0175±0.0039 TS Longitudinal
o
C65-WEB
80
1.8
CO
60
^
C
(U
c
CD
o
- 40
§^1.6
Core column
C8C1 F,=34 ksi
1.4
_l
I
L.
_l
-2
19:47:55
•
Wednesday Febmary
I
L
0
log.io strain rate, 1/s
09,
2005
Figure A-73. Strain rate sensitivity for Fy = 36 ksi core wide-flange C80 longitudinal.
NISTNCSTAR
1-3D.
WTC
Investigation
231
Appendix A
2.0
-
O
100
m=0.0265±0.0014 F^, Longitudinal
m=0. 0200+0. 0058 TS Longitudinai
80
CO
1.8
60
^
c
C
CO
CD
(f)
8^1.6
40
Core column
HH_web F =52
J
1.4
I
L.
I
-2
19:48:02
-
Wednesday February
09,
I
Ul
I
ksi
I
0
log 10 strain rate, 1/s
2006
Figure A-74. Strain rate sensitivity for Fy = 42 ksi core wide-flange
HH1-WEB
longitudinal.
"1
O
2.0
'
———
'
F,,
Tfe
'
1
1
r
n
1
r
100
Transverse
Transverse
80
CO
1.8
CO
60
I—
^
O)
c
CD
i_
05
CO
CD
c
CO
8^1.6
1.4
40
I
Core column
HH_web F V =52
L
__]
-4
-2
-
Wednesday February
09,
I
I
ksi
I
0
log.,0 strain rate,
19:48 02
'
1/s
2005
Figure A-75. Strain rate sensitivity for Fy = 42 ksi core wide-flange
232
HH1-WEB
NISTNCSTAR
1-3D,
transverse.
WTC Investigation
Data Tables and Supplemental Figures
1
1
1
1
1
1
1
1
1
2.0
-
O
—I
1
1
1
1
- 100
m=0.0239±0.0038 F^, Longitudinal
m=0.01 65+0.001 2 TS Longitudinal
-
80
CO
c
60
o
-
(U
^
C5)
C
CD
o
/S
'
40
1.6
o
1.4
1
1
1
1
-4
1
1
1
L
Core column
B^152-2 Fy=38
-2
0
k^i
2
log^o strain rate, 1/s
19:47:57
-
Wednesday February
09.
2005
Figure A-76. Strain rate sensitivity for Fy = 36 ksi box column B6152-FL longitudinal.
\
2.0
O
I
i
100
m=0.0396±0.0009 F Transverse
m=0.0196±0.001 7 TS Transverse
80
CO
60
^
c
C
w
CD
o
8^1.6
40
Core column
B^152-2 Fy=38
J
1.4
I
-4
J
L.
-2
I
k^i
L
0
log.,0 strain rate, 1/s
19:47:58
•
Wednesday February
09,
2005
Figure A-77. Strain rate sensitivity for Fy = 36 ksi box column B6152-FL transverse.
NISTNCSTAR
1-3D.
WTC
Investigation
233
Appendix A
50
-r
*
A
I
+
T
T-
1
'
\
r-
Tolal Elongation
F=KX)
k-«i e.=( 1.98+0.23) MIOB-C^B-RlF,= IOO ksi e|=(1.24±0.07) ClO-ClM-LF
F;= 75 k,sie,=(l. 27+0.1 1)C14-SP
El =e„+e|log,(dE/dt)
40
c
o
20
10
0
-2
log,() Strain rate, 1/s
50
T
F -iOO
F,= 00
1
ksi
k.si
1
r
c.=(0.86±0.18) M10B-C3B-RF
e,=( 1 .20±0.0 ) CI 0-C M-LF
i
1
40
.2
-"^0
c
O
20
perimeter steeks longitudinal orientation
10
-2
0
logjo strain rate, 1/s
n.HM)!'
-
Wediicsduy December
15, 2(1114
Figure A-78. Dependence of total elongation, £/,, on strain rate for high-strength
perimeter columns.
234
NISTNCSTAR
1-3D,
WTC Investigation
Data Tables and Supplemental Figures
-1
p
1
I
t
I
I
t
\
I
O
=50
ksi
F,=60
F,=65
f;=75
ksi
F,
A
0
•
ksi
ksi
Reduction of area
B-RF
ri=(1.92±0.21) N8-C1B-RF
r,=(0.85±0.11 N99-C3T-RF
r,=(0.22±0.18) C22-C2M-LF
r,=(0.75±0.33) M26-C1
ROA
= ro+r,log^(de/dt)
f
-2
0
loQio Strain rate, 1/s
T
1
1
1
1
\
1
r
T
r
1
60
<
o
40
F„=50
F'=60
F.=65
F'=75
ksi
r,=(-1.10±0.30)
ksi
N8-C1B-RF
ksi
r,=(-2.64±0.85)
.9310.57)
ksi r,=(-1
M26-C1B-RF
N99-C3T-RF
C22-C2M-LF
perimeter steels longitudinal orientation
_l
-2
I
1
I
L.
I
t
I
I
0
loQio Strain rate, 1/s
18:40.29
-
Thursday October 28 2004
Figure A-79. Dependence of total elongation, £/,, on strain rate for low-strength
perimeter columns.
NISTNCSTAR
1-3D.
WTC
Investigation
235
Appendix A
50
T
1
1
1
1
1
1
P =36
0
•
r,,=:36 ksi e.=--{+0.5Q±" .37)
F =36
F,=42
ksi
ksi
ksi
—
p-
1
1
A
'
'
'
\
Elongation
e,=(+0.83±0.27) B61 52-2
C65-weD
El,
= eo+e,logp(dE/dt)
ei=(+0.62±0.56) C65-flange
e, ={-0.1 9+0.02) HH-web
40
A
A
.2 30
D5
C
o
CD
20
all
Flat-HSR
specimens
1"
with t=0.0625
in.
A"^/Lo = 0.125
core steels longitudinal orientation
J
10
L.
I
J
I
I
I
I
I
I
I
I
I
0
-2
log^o strain rate, 1/s
14 19 20
-
Friday
December
17.
2004
Figure A-80. Dependence of total elongation,
-1
—
-
-
1
1
1
1
r 1
Elt,
on
—
strain rate for core
columns.
1
1
1
1
1
True stress
True strain rate
r
test
1
WTC
1200
\
1
1
1
434 C22-C2M-FL3
1
Col 157 93-96
4000
1000
03
3000
D.
cd'
800
CO
c/)
00
600
2000
CD
13
§CO
CD
400
I-
- 1000
200
J
I
I
I
0.0
I
0.1
I
0
L.
0.2
True
1
7:21
25
.
Monday December
20,
0.3
0.4
strain
2004
Figure A-81. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
236
NISTNCSTAR
1-3D.
WTC
Investigation
Data Tables and Supplemental Figures
test
4k
WTC
1200 -
1
C22-C2M-FL3
Col 157 93-96
4000
1000
CD
3000
D.
800 -
c5
(0
-—
600
w
2000
2
CO
CD
13
0)
400 f
1000
200
0.0
True
17:21:26
Monday December
-
20,
0.4
0.2
0.1
strain
2004
Figure A-82. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
T
1
1200 -
1
1
1
1
1
r
1
True stress
True strain rate
test
I
I
I
I
436 C22-C2M-FL3
WTC
1
Col 157 93-96
4000
3000
—
•4
CD
2000
2
CO
o
3
1000
_L
0.0
0.1
0.2
True
17:21:28
-
Monday December
20,
0.3
0.4
strain
2004
Figure A-83. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
NISTNCSTAR
1-3D.
WTC
Investigation
237
Appendix A
—
=
1
1
1
1
1
1
1
1
1
1
r
1
1
1
1
1
1
1
1
1
True stress
True strain rate
test
4J/ \jdd-'^dm-r\-6
WTC
1200
'
1
Col 157 93-96
4000
1000 -
C5
3000
CL
800
w
•4—
(/)
CO
/^^
-
0)
-
/
600
2000
%
\
13
CD
3
-/
400 —
I-
-1
1000
200 \
J
\
\
1
1
t
1
0.0
1
1
1
1
1
1
True
17:21:28
•
Monday December
20,
1
1
1
1
1
1
1
1
1
1
0.4
0.3
0.2
0.1
strain
2004
Figure A-84. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
T
1
1
1
1
1
1
— —T
r
1
True stress
True strain rate
test
WTC
1200
^
\
^
s
438 C22-C2M-FL3
1
f'
Col 157 93-96
4000
1000
CO
CT3
3000
CL
800
"cD
CO
0)
00
600
2000
2
CO
CD
3
CD
3
400
1000
200
J
0
0.0
0.1
I
1
I
True
17,21 29
-
I
I
0.2
LIO
i_
0.3
0.4
strain
Monday December 20 2004
Figure A-85. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
238
NISTNCSTAR
1-3D.
WTC Investigation
Data Tables and Supplemental Figures
0.0
0.2
0.1
True
Monday December
17:21:30
20.
0.4
0.3
strain
2004
Figure A-86. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
T
1
1
1
1
1
1
1
p
1
True stress
True strain rate
1200 -
test
440 C22-C2M-FL3
WTC
1
"
Col 157 93-96
4000
1000
CO
3000
Q_
0)
800 CO
(0
<D
—
.*
0)
600
2000
0)
13
2
O
Z3
400 1000
200
J
I
I
I
'
0.1
I
I
I
I
I
True
17:21:31
-
Monday Decen*er
20.
I
I
0.2
I
I
I
0.3
L
0.4
strain
2004
Figure A-87. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
NISTNCSTAR
1-3D.
WTC
Investigation
239
Appendix A
1
1
1
1
1
-n
r
1
1
True stress
True strain rate
1200
——
I
1
I
441 C22-C2M-FL3
WTC
•
>
1
test
1
f"
Col 157 93-96
4000
1000
CO
3000
Cl
800
CO
C3
600
2000
O
3
S
(J)
3
400
- 1000
r//
200
I
I
I
I
0.0
I
0.1
True
17:21:31
-
Monday December
20.
I
0.2
1
I
0.3
0.4
strain
2004
Figure A-88. Kolsky-test stress-strain behavior for Fy = 75 ksi perimeter column
C22-C2M.
1200 -
1000
03
Q_
800 CO
0)
CO
600 -
CD
Z3
400
-J
200
0.2
True
17:21:32
-
Monday December
20.
strain
2004
Figure A-89. Kolsky-test stress-strain behavior for Fy = 60 ksi perimeter column N8-C3M.
240
NISTNCSTAR
1-3D.
WTC
Investigation
Data Tables and Supplemental Figures
test
N8-C3M-FR2
4^1
WTC
1200 -
1
Col 141 97-100
4000
1000
(0
3000
CL
800 -
c5
600
00
2000
2
CO
400
I-
1000
200
0.0
0.2
0.1
True
17:21:32
-
Monday December
20.
strain
2004
Figure A-90. Kolsky-test stress-strain behavior for Fy = 60 ksi perimeter column N8-C3M.
I
I
—
1200 -
I
I
\
—
I
—
p
I
T
>
>
'
>
452 N8-C3M-FR2
WTC 1 Col 141 97-100
True stress
True strain rate
r"
test
'
4000
1000
CO
CO
3000
Bc5
800
(0
600
2000
o
^m
CD
15
400
- 1000
200
I
I
I
1
0.0
I
0.1
_l
1
True
17:21 :33
•
Monday December
20,
1
0.2
Lio
L.
0.3
0.4
strain
2004
Kolsky-test stress-strain behavior for Fy = 60 ksi perimeter column N8-C3M.
Figure A-91
.
NISTNCSTAR
1-3D,
WTC Investigation
241
Appendix A
—
r
1
1
1
1
1
n
r
I
True stress
True strain rate
—I
I
1
1
453 N8-C3M-FR2
WTC
1200
1—
1
test
1
Col 141 97-100
4000
1000
CO
- 3000
CL
(D
4—
800
CO
CO
600
2000
2
(0
0)
Zi
0)
400
1000
200
/
0
J
I
I
I
0.0
!_)
I
I
I
I
I
True
1
7:21 :34
-
Monday December
20.
I
I
I
— —U
J
L.
0.2
0.1
I
0
0.4
0.3
strain
2004
Figure A-92. Kolsky-test stress-strain behavior for Fy - 60 ksi perimeter column N8-C3M.
^
I
I
I
r
T
I
p-T
r
\
1
\
I
454 N8-C3M-FR2
WTC 1 Col 141 97-100
True stress
True strain rate
1200 -
r
test
4000
1000
V)
3000
0)
800
o5
C/3
Q)
CO
600
2000
0)
13
5CO
CD
400
I
1000
I
200
J
I
0.0
I
0.1
I
I
I
I
I
True
1
7:21 35
-
Monday December
20.
I
I
0.2
I
I
I
0.3
I
I
I
L
Ll
0
0.4
strain
2004
Figure A-93. Kolsky-test stress-strain behavior for Fy = 60 ksi perimeter column N8-C3M.
242
NISTNCSTAR
1-3D,
WTC Investigation
Data Tables and Supplemental Figures
Figure A-94. Kolsky-test stress-strain behavior for Fy = 60 ksi perimeter column N8-C3M.
Figure A-95. Kolsky-test stress-strain behavior for Fy = 60 ksi perimeter column N8-C3M.
NISTNCSTAR
1-3D.
WTC
Investigation
243
Appendix A
T
1
1
1
1
1
1
1
1
1
1
1200 -
-1
r
1
1
1
I
I
457 N8-C3M-FR2
WTC 1 Col 141 97-100
True stress
True strain rate
test
4000
/ \
1000
n3
3000
Cl
B
800
c/)
CO
•4—
W
600
2000
0)
^w
CD
3
400
1000
200
J
0.0
I
I
I
1
0.1
I
I
I
I
I
True
17:21 43
•
Monday December
20,
I
I
0.2
I
I
I
0.3
I
I
1
L.
0.4
strain
2004
Figure A-96. Kolsky-test stress-strain behavior for Fy - 60 ksi perimeter coiumn N8-C3M.
1200
4000
1000 -
- 3000
Q_
0)
800
to
(0
600 -
- 2000
0)
13
5
0)
Z5
400
- 1000
200 -
0.2
True
1
7.21 44
-
Monday December
20,
strain
2004
Figure A-97. Kolsky-test stress-strain behavior for Fy = 60 ksi perimeter column N8-C3M.
244
NISTNCSTAR
1-3D,
WTC
Investigation
•
'
Data Tables and Supplemental Figures
~i
1
1
1
1
1
-T
r
T
1
r—
1
324 A 325 bolt
unknown location
True stress
True strain rate
test
1200
4000
1000
(0
3000
Q.
0)
800
CO
(/}
CO
CD
c
-—
"cO
w
600
2000
7
—
•4
(0
0)
0
3
400
1000
200
1111
0
0.0
±
0.1
0.2
True
17:21:46
-
Monday December
20.
J
I
I
0.0
0.1
strain
0.2
True
7:21 49
•
Monday December
20.
1-3D.
WTC
A
325
bolt.
0.4
0.3
strain
2004
Figure A-99. Kolsky-test stress-strain behavior for
NISTNCSTAR
0.4
2004
Figure A-98. Kolsky-test stress-strain behavior for
1
L
0.3
Investigation
A
325
bolt.
245
Appendix A
T
1
1
1
1
r
1
1
1
1
1
1200 -
———
I
1
r
True stress
True strain rate
>
test
3^6 A 325
unknown
bolt
location
4000
1000
3000
CL
CD
800
CO
C/3
if}
CD
1
600
2000
-
CD
(
5CO
. /
3
CD
=3
400
r
/ \
I
1000
200
0
J
I
I
L
0.0
I
I
I
I
I
i
True
17:21 50
-
Monday December
20.
I
I
I
I
0.2
0.1
I
I
I
I
0.4
strain
2004
Figure A-100. Kolsky-test stress-strain behavior for
T
1
1
l_
0.3
I
1
A 325
bolt.
r
True stress
True strain rate
test
327 A 325
unknown
1200
bolt
location
4000
1000 CO
3000
Q.
800
"co
CO
CD
V)
600 -
H
2000
CD
tm
CD
400
I-
- 1000
200 L
I
0.0
0.1
0.2
True
17:21:52
-
Monday December
Figure A-1 01
246
.
20,
I
I
I
0.4
0.3
strain
2004
Kolsky-test stress-strain behavior for
A
325
NISTNCSTAR
bolt.
1-3D.
WTC
Investigation
•
Data Tables and Supplemental Figures
1
1
1
1
1
1
1
1
1
1
r
1
1
1
1
1
1
1
1
1
True stress
True strain rate
"
test
328 A 325
unknown
1200
bolt
location
4000
-
1000
- 3000
Q.
800
/
CO
(0
o
w
^—
600 -
400
tv^
c
2000
I
'cO
CO
-
\
i
-
!
\\
CD
=3
CO
I
/
CD
\
i_
'
I-
1000
200 r
I
0
1
1
1
1
0.0
1
0.1
1
t
1
1
1
True
1
NISTNCSTAR
7:21 :54
-
Monday December
20.
1
1
0.2
1
1
1
1
1
1
0.3
1
1
0.4
strain
2004
Figure A-102. Kolsky-test stress-strain behavior for
A 325
bolt.
Figure A-103. Kolsky-test stress-strain behavior for
A 325
bolt.
1-3D.
WTC
Investigation
247
Appendix A
A.7
SUPPLEMENTAL FIGURES FOR ELEVATED-TEMPERATURE
PROPERTIES
Figures
A- 104
through A-1 12 summarize the high-temperature stress-strain behavior of all steels tested.
1
00
1
1
1
1
1
1
1
1
1
1
r
1
1
1
I
annotations refer to individual specimen numbers
0
'
I
I
I
1
I
0.0
I
I
I
I
I
1
\
I
0.2
0.1
I
I
I
I
I
I
0.4
0.3
Engineering Strain
10:57:55
-
Friday Octobei 08,
2004
Figure A-1 04. Perimeter column Fy = 60 ksl N8-C1B-LF.
1
00
——————————————I— — ——
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
t
I
0.0
0.1
0.2
0.3
0.4
Engineering Strain
12:00 18
-
Frrday Oclober29.
2004
Figure A-1 05. Perimeter column Fy = 60 ksi C40-C2M-IW.
248
NISTNCSTAR
1-3D,
WTC
Investigation
r
Data Tables and Supplemental Figures
1
—— — — —— — — — —
I
I
I
I
I
I
I
0.00
I
I
!
I
1
I
I
0.05
I
I
I
I
I
0.10
I
I
I
0.15
I
I
I
I
I
I
I
I
I
I
0.20
0.25
0.30
0.35
Engineering Strain
1733:31
-
Fnday October 08. 2004
Figure A-106. Perimeter column Fy = 100 ksi C10-C1M-FL.
'
T
Q
I
1
1
1
1
1
1
1
I
I
I
0.0
I
I
0.1
I
I
1
1
I
0.2
1
1
1
1
1
1
I
I
I
I
I
0.3
1
1
I
I
I
I
I
r
1__L
0.4
Engineering Strain
12:03:37
-
Fnday October 29. 2004
Figure A-107. Core wide-flange column Fy - 36 ksi C65-FL longitudinal.
NISTNCSTAR
1-3D.
WTC
Investigation
249
Appendix A
I
I
I
0.0
I
I
,,
I
I
I
,
I
I
I
I
I
I
I
I
I
I
0.4
0.3
0.2
0.1
Engineering Strain
12:04:46
Friday October 29.
2004
Figure A-108. Core wide-flange column Fy = 36 ksi C80.
HH-FL-1
WTC1 605
0
I
I
I
I
I
0.0
I
0.1
I
I
I
I
I
I
I
0.2
I
I
I
I
I
I
98-101
I
I
I
I
I
0.3
0.4
'
0.5
Engineering Strain
1 1
:51 :49
-
Sunday September
12,
2004
Figure A-109. Core wide-flange column Fy = 42 ksi HH.
250
NISTNCSTAR
1-3D,
WTC
Investigation
Data Tables and Supplemental Figures
1
1
1
1
r
1
1
#7w4; 20 °C
#2w7; 400 °C
C
#1n3: 500
#4n3: 600 °C
600 °C
#6w1; 650 °C
CI 32
2
_/
I
0.0
I
I
I
I
I
X
in.
1
.5 in. truss
I
I
_I
0.2
0.1
angle
L
I
I
L
I
0.4
0.3
Engineering Strain
12:05:57
-
Fnday October
Figure A-1
1 0.
29.
2004
Truss 2
in.x1 .5 in.x0.25 in.
n
I
!
I
I
I
angle Fy - 50 ksi C1 32.
1
n
1
#2-5w; 300
r
1
C
#1-6n; 20
#3-6;
°C
400 °C
C53
3
X
in.
2
in,
truss angle
#1-3w; 600
C
#2-4w; 650
0
I
I
I
I
I
0.00
I
0.05
I
L.
J
0.10
I
I
I
I
I
I
C
I
I
0.20
0.15
0.25
Engineering Strain
Figure A-1 1 1
NISTNCSTAR
1-3D.
WTC
.
Investigation
Truss 3 in.x2 in.x0.37
in.
angle Fy - 36
ksi.
251
Appendix A
—— — —
1
I
1
I
"1
1
I
I
I
I
I
r~T
I
I
I
r
I
#1254-1; 20 °C
#1255Y-1; 20 °C
C
#1254-2: 400
-
#1255Y-2: 400 °C
1254 prepared by
AWS A5.20-69
AWS A5.1-69
255Y prepared by
#1255Y-3; 650 °C
#1254-3; 650 °C
J
0.0
0.1
I
I
I
I
0.3
0.2
I
I
I
I
I
I
I
i_
0.4
0.5
Engineering Strain
12:08:40
-
Friday October 29,
2004
Figure A-112. Contemporary
252
all
weld metal specimens.
NISTNCSTAR
1-3D,
WTC Investigation
Appendix B
Effects of Deformation of Wide-Flange Core Columns on
Measured Yield Strength
INTRODUCTION
B.1
Several of the stress-strain curves of specimens taken from wide-flange core columns differ significantly
between the webs and the flanges. In
particular, the stress-strain cur\'es
of specimens from the flanges of
C-65, C-71. and C-155 and the webs of C-65 do not exhibit yield point or yield point elongation behavior.
The
yield strengths, YS, of specimens from the flanges of C-71 and
from the corresponding webs. Both
Although National
Institute
deformed material
in
was
that
effects are evidence
C-155
much
larger than those
of Standards and Technology (NIST) made every effort to sample the
core column, the condition of the core columns
made
deformed. Measurements of the shapes of the columns
slightly
are also
of significant prior deformation.
used to estimate the level of strain
the Investigation. This appendix
in tensile
it
in their as-recovered state
can be
specimens harvested from those columns and characterized
summarizes
in
this analysis.
analysis
B.2
Figure B-1 defines the variables that this analysis uses. In a typical wide-flange column, the
inertia
first
least-
necessary to take material
about the
.v-x
axis
is
about three times as large as the
moment of
axis. To a
moment of inertia about the v -v
moment of inertia; the major
approximation, a column will buckle about the axis with the lower
deformation will be about the v-v axis. The deformation of a wide-flange shape bent about the x-x axis
called camber, while the deformation about the>'-v
If the
column bends around
the y-y axis, the strain,
f,,
is
is
called sweep, see Fig. B-2.
£,.,
at
any point
in the flange is
=-
B-8
R
where
the
is
beam
the distance
is
bent
to.
from the neutral axisj-v
=6
in.
from the neutral
In practice,
/?,,
it
is
and sweep,
,
the centerline of the web.
axis, in contrast to
R
less strain in the
is
the radius of curvature
web of the column
Specimens from the web originate only
a
specimens from the flanges, which originate up to
axis.
difficult to
/?,
down
from the neutral
fraction of an inch
specimen position, and
Bending about the y-y axis produces much
because the neutral axis runs
/,
to the
measure the radius of curvature, R. Instead,
over a length
/,
of the column
in the
it
is
easier to
measure the camber,
region where the tensile specimens originated. The
radius of curvature, R, and the sweep, hy, are related geometrically:
NISTNCSTAR
1-3D.
WTC
Investigation
253
Appendix B
h +
—
R=Z
B-9
4.
Vi.
much larger than the thickness of the piece.
be much larger than the camber, because the
Equation B-2 assumes that the radius of curvature, R,
sweep of a buckled column
practice, the
moment of inertia about
the x-x axis
is
will usually
In
is
smaller.
RESULTS
B.3
Figures
B-3
to
B-6
montages of the core coluinns C-30, C-65, C-71, and C-155
are photo
and
the overall condition
specific origins of tensile test specimens. Table
B-1 summarizes
dimensions of these four columns. Table B-2 summarizes the measured sweeps,
webs and
The sweep of column C-30
their resulting strains.
is
that describe
/?,
,
the important
of the flanges and
indistinguishable from zero.
The curvatures
of columns C-71 and C-155 were verified by measuring the sweep of the flange on both sides of the web.
As expected,
The
it
was
positive on one side and negative and of approximately equal magnitude on the other.
sweep of the column, +/-0.02
large uncertainty in the
edge of the flanges, which were presumably damaged
in.,
arises
from the rather irregular shape of the
in the collapse
and subsequent recovery.
Table B-3 summarizes the measured cambers of the four core columns. The cambers of the flanges of
columns C-30, C-71, and C-155 are indistinguishable from
indistinguishable
from
deformation of the
zero,
camber
web of C-65.
shape of its web, Fig. B-4,
is
Its
is
Although the sweep of column C-65 was
zero.
measurable. Fig. B-^. Table B-3 also contains an entry for the
shape does not correspond to either camber or sweep. Instead, the
consistent with the entire
web having
buckled.
DISCUSSION
B.4
From
its
the
measured sweep,
/?,
,
it is
Equations B-1 and B-2. Table
strain in the flange test
possible to estimate the strain,
B-2 summarizes
specimens
is
about
1
this
8,
,,
in the flange
specimens using
infonnation in the column labeled
originated, the strain
is
webs and
flanges of C-71 and
same magnitude
tensile in
C-155
is
254
test
specimens
all
four columns.
certainly sufficient to
remove
as the typical yield point elongation.
The
The
strain estimated
from the sweep of the
the flange yield point behavior,
flange tests of C-71 and
strain calculated
and
web
is
the
C-155 do not exhibit
from the sweep of the
Column C-30, which has no measurable sweep or camber, exhibits yield point behavior
flange and
The
specimens from both columns. Table B-2 also summarizes the yield
flanges of
any obvious yield point elongation behavior, consistent with the
flange.
shape."'
percent in columns C-71 and C-155, where the sweeps could
be measured. Based on the sign of the sweep on the side of the flange from which the
point behavior of the
"£,.,
in
both
tests.
NISTNCSTAR
1-3D,
WTC
Investigation
Effects of Deformation
The sweeps reported
ASTM A
370. which
in
Table B-2 are roughly five times larger than the
maximum sweep
permissible in
is
/
h""''
=0.125
B-10
in.
V
The maximum allowable sweep and camber
the core columns,
A
370
is
in.
smaller than the smallest sweep measurable on
due to their short length and poor edge condition.
Both flange and web
test
specimens for column C-65 originated
deformation for these specimens
strain
in
120
is
depends sensitively on the exact location of the specimen
types of flange specimens for
roughly
at
the
1
in
deformed material. Because the
predominantly camber, rather than sweep, the magnitude of the pre-
column C-65. Some were 0.25
in.
4-depth point of the flange. The others were 0.5
in the original billet.
There were two
diameter rounds whose centerlines lay
in.
diameter rounds whose centerlines lay
roughly on the mid-depth of the flange. Because of the camber of the flange, both types of specimens
have non-uniform pre-strain across
their widths.
The deformation
is
probably sufficient to remove any
yield point behavior, however.
It is
also possible, neglecting
the flange specimens of C-71
This
is
a coarse
any contribution of the Bauschinger
and C- 155 by assuming
that the
effect, to
web
roughly estimate the strain in
represents the
undeformed behavior.
assumption because the flanges typically lower yield strength. The flange strains
predicted from the
web
stress-strain
behavior are approximately 4 percent and 2 percent for C-71 and
C-155 respectively.
SUMMARY
B.5
The magnitudes of the prior
strains estimated
from the
measured sweep of core columns C-71 and C-155.
and lack of yield point behavior
the column.
tensile
in the flanges
The deformation of column C-65
specimens cut from
NISTNCSTAR
1-3D.
WTC
It is
tensile data are similar to those estimated
from the
likely, therefore, that the elevated yield strength
of these two columns arises from the prior deformation of
is
sufficient to
remove any
yield point behavior
from the
it.
Investigation
255
Appendix B
3-1. Locations
and dimension informaltion
for wide-f
ange core co
d
Shape
Specimen
Location
WTC 2
14WF237
C-30
A
(in.)
(in.)
(in.)
(in.)
(in.)
16.12
15.91
1.09
1.75
5.0
13.88
12.52
0.91
1.49
2.3
14.38
12.67
1.06
1.74
4.0
13.88
12.52
0.91
1.49
4.0
Col. 1008
Fl.
WTC
12WF161
C-65
104-106
1
Col. 904
Fl.
WTC
12WF190
C-71
86-89
1
Col. 904
Fl.
WTC
12WF161
C-155
79-80
1
Col. 904
Fl.
83-86
Note: See Fig. B-1 for definitions of symbols.
Table B-2. Stress-strain behavior and sweep of core co umns.
Web
£,,,,
£,,
K
(ksi)
(ksi)
Shape"
Test"
(in.)
(in.)
(in.)
(in.)
41.9 (YP)
39.6 (YP)
<0.004'
n/a
<0.02
<0.02
15
> 1,400'
Specimen
C-30
C-65
31.1
(NYP)
Flange
(NYP)
<0.0004'
n/a
<0.03
n/d
36
>5,400'
0.09
0.09
16
380
0.13
0.11
18
320
33.2 (YP)
53.4
(NYP)
+0.011
0.04
C-155
42.2 (YP)
50.9
(NYP)
+0.012
0.02
a.
Estimated from sweep of flange,
b.
Estimated from stress-strain behavior of flange, relative to the web.
c.
Uncertainties in
d.
/ is
e.
Estimated from
/?,
/?,
,
and location of specimens,
are approximately +/-0.02
the length of the flange over
n/a, not
R
32.4
C-71
Key:
Web
Flange
Fy
/,
using Eqs. B-1 and B-2.
in.
which the sweep
is
measured, Fig. B-1
minimum measurable sweep,
applicable; n/d, not determined; NYP, no
13.
/?,
yield point behavior in stress-strain curve;
YP,
yield point behavior in stress-strain cur\'e.
256
NISTNCSTAR
1-3D,
WTC Investigation
Effects of Deformation
Table B-3. Stress-strain behavior and camber and deformation of core columns.
Web
Specimen
Flange
Fy
(ksi)
(ksi)
(in.)
(in.)
flange
<0.02
15
'"A
C-30
41.9 (YP)
C-65
31.1
(NYP) 32.4 (NYP)
flange
0.17
36
C-65
31.1
(NYP) 32.4 (NYP)
web
0.38
9
C-71
33.2 (YP)
53.4
(NYP)
flange
<0.015
16
C-155
42.2 (YP)
50.9
(NYP)
flange
<0.015
18
a.
Uncertainties in
b.
/ is
/?,
39.6 (YP)
are approximately +/-0.02
the length of the flange over
Key: NYP, no yield point behavior
behavior
NISTNCSTAR
Flange
Measurement
Type
1-3D,
in.
which the camber
is
measured.
in stress-strain ciir\'e;
YP. yield point
in stress-strain curve.
WTC Investigation
257
Appendix B
Source: NIST.
Figure B-3. Montage of specimen C-30 that shows the overall state of the column and the
region that provided the test specimens.
258
NISTNCSTAR
1-3D.
WTC Investigation
Effects of Deformation
Source: NIST.
Figure B-4. Montage of specimen C-65 that shows the overall state of the column and
the region that provided the test specimens.
NISTNCSTAR
1-3D,
WTC
Investigation
259
Appendix B
1
6
in
C71-1-VV1
4
.
3
in.
in.
C71-1-F1
C71-1-F1
tension specimens originated
specimen C71-1-F2, which came from
C71-1-F2
All
flange (hidden by
web above)
with
tensile strain. Note that the curvature
of the specimen area is opposite to the
general curvature of column.
C71-1-.W1
Source: NIST.
Figure B-5. Montage of specimen C-71 that shows the overall state of the column and the
region that provided the test specimens.
260
NISTNCSTAR
1-3D,
WTC
Investigation
Effects of Deformation
Source: NIST.
Figure B-6. Montage of specimen C-1 55 that shows the overall state of the column and
the region that provided the test specimens.
NISTNCSTAR
1-3D,
WTC
Investigation
261
Appendix B
This page intentionally
262
left
blank.
NISTNCSTAR
1-3D,
WTC
Investigation
.
Appendix C
Provisional Analysis of High-Rate Data
Early in the Investigation, before testing was complete,
other Investigation
Team members modeling
was necessary
it
to provide high-strain-rate data to
The accelerated
the aircraft impact on the building.
schedule required that the model for the data be developed on a provisional set of the
provisional data set includes the six steels listed in Table C-1.
The reported values
steels.
in
This
Appendix
A differ
shghtly from the provisional values used in the provisional analysis because the provisional analysis did
not use the
For use
ESIS procedure
for data reduction.
element model, the data had to be expressed
in the finite
in
terms of the Cowper-Symonds
equation:
—
6
C-1
=i-t-
The Cowper-Symonds equation describes the increase in flow stress with strain
parameters. C and p. In the Cowper-Symonds equation s is the strain rate, a is
or other stress) and ao
The present
is
the reference stress. Usually, Gq
is
the stress
measured
rate using
two
the stress (yield, tensile,
in a low-rate tensile test.
analysis uses the yield strength.
Equation C-1 can be rewritten as
^^
'
=s =
0
s
^
C-2
C/
Taking the natural logarithm of each side produces an equation linear
in the
Cowper-Symonds
parameters:
log^
S =
- log, ^ - -log, C
The value of the
stress at the
cannot be included
To avoid
this
in a fit
problem.
lowest strain
fit
a line to the cr
rate test data. Plotting this semi-log
this
curve yields the
In these
fits
rate, Oo, fixes the overall strain rate
of Equation C-3, because
NIST
fit
C-3
P
P
at
a=
- log, £
on log-log axes
Oo,
data, as
is
commonly done
results in a cur\'e.
Cowper-Symonds parameters p and
behavior, but this value
S =0
A
linear
fit
in analysis
of high-
of Equation C-3 to
C.
Cowper-Symonds exponent,/?, was constant among the steels,
yield strength. The increase in C with increasing yield strength, F,.,
for the provisional data, the
but the pre-factor
C
increased with the
can be represented with a second order polynomial of the natural logarithm of C:
NISTNCSTAR
1-3D.
WTC
Investigation
263
Appendix C
log,
where F,
is
the yield strength
Cowper-Symonds model,
measured
C = C„+C,F,+CF;
C-4
in the low-rate tensile test in units
after incorporating the yield strength
(7
of ksi. The provisional
dependence of the parameter
C is
= +
C-5
1
Table C-2 summarizes the values of the parameters in the provisional model, Eq. C-5.
The predictions of Eq. C-5
were reasonable
to the individual provisional data sets for the six steels
in that
the yield stress increases with increasing strain rate. Furthermore, predicting the behavior of low-strength
steels, outside the strength
range of the original data
set,
using Eq.
C-5 does
not produce anomalous
behavior.
By
analyzing the
data
set,
it is
full
data set collected as part of the Investigation in the
same manner
possible to assess the quality of the predictions of the provisional model.
of the provisional data, the exponent p for the
The nomializing constant
C also
full
as the provisional
As
in the analysis
data set does not vary between the steels:
p = 4.4053.
increases with increasing yield strength; a second-order polynomial
sufficient to represent the behavior. Table
C-2 summarizes the parameters
in the
is
Cowper-Symonds model
for the full data set of 13 steels.
The model based on
the full data set predicts the behavior of the six steels in the provisional data set about
as well as the provisional
model does. Of these
cases,
where the goodness-of-fit
model
fits
is
taken as the
six steels, the full
model
sum of squares of the
fits
more
residuals.
accurately in half the
As
expected, the
full
the data of the steels not in the provisional set better than does the provisional model. In the
cases of the high-strength steels, the differences are not substantial, however. Figures C-1 through
illustrate the data
and the predictions of both models for high and intermediate strength
The provisional model tends
steels.
Figures
steels that
first,
264
to overpredict the strain-rate sensitivity
C-10 through C-13 show
were not included
the predictions of the
in the provisional data set.
the reduced fidelity for the core
of the low-strength, core column
model
Because the
columns should not be a
C-13
steels.
for the low-strength, core-column
aircraft struck the perimeter
columns
significant problem.
NISTNCSTAR
1-3D,
WTC Investigation
Provisional Analysis of High-Rate Data
Table C-1. Specimens and orientations used
in
the provisional analysis to determine
Fy
(ksi)
Specimen
Orientation
50
M26-C1B
L&T
60
N8-C1B
L
65
N99-C3T
L&T
75
C22-C2M
L
100
M10B-C3B
L&T
100
ClO-ClM
L&T
Key: L,
longitudinal; T, transverse.
Table C-2. Comparison of provisional and
final
Cowper-Symonds parameters,
Eq. C-5,
for the high-strain rate data.
Parameters in Equation C—4
Data Set
# of Steels
P
G,
c,
Complete
13
4.4053
-4.7374
0.3614
-0.001707
Pro\isional
6
6.7824
-17.391
0.74512
-0.0035253
(Table C-1)
NISTNCSTAR
1-3D.
WTC Investigation
265
Appendix C
Figure C-1. Predictions of the provisional and
column
1
full
C10-C1M1-FL which
steel
Cowper-Symonds model
is in
,
—— — — —— — — — ——— —
I
I
I
I
I
I
I
I
I
I
I
for perimeter
the provisional data set.
1
I
—— — —— — ——
I
I
I
I
I
I
I
provisional set
set
full
1.3
O
o
1.2
o
o o
N99-C3M F =73.6
ksi
1-148-99/102
1.0
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
400
300
200
100
0
I
L
500
Strain rate, 1/s
16:56:29
-
Tuesday December
14.
2004
Figure C-2. Predictions of the provisional and
column
266
steel
C10-C1M1-IW which
,
full
is
Cowper-Symonds model
not
in
for perimeter
the provisional data set.
NISTNCSTAR
1-3D,
WTC Investigation
r
r
r
Provisional Analysis of High-Rate Data
~i
——— —— — — — —
I
I
I
I
I
I
I
———
1
I
I
I
I
I
I
I
c
I
"I
provisional set
set
full
B6152-2-FL F =40.2
ksi
1-803-33/36
1.0
^
_L
0
100
J
I
I
J
L.
I
200
I
I
I
400
300
L.
1
i
500
Strain rate, 1/s
15:56:12
-
Tuesday December
14.
2004
Figure C-3. Predictions of the provisional and
column
full
M10b-C3B1-RF, which
steel
Cowper-Symonds model
is in
for perimeter
the provisional data set.
—— ————— — — — — — — — — —
1
1.20
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
provisional set
set
full
O
1.15
O
o
D
D
1.10
1.05
M10B-C3B F =112.6
2-206-83/86
1.00
J
1
I
I
I
100
0
I
1
I
I
I
I
I
I
200
I
I
I
I
I
300
I
I
400
I
I
1
L
ksi
I
,
500
Strain rate, 1/s
15:55:36
-
Tuesday December
14.
2004
Figure C-4. Predictions of the provisional and
column
NISTNCSTAR
1-3D,
WTC
steel
C22-C2M, which
Investigation
full
is in
Cowper-Symonds model
for perimeter
the provisional data set.
267
Appendix C
—— — — — ——— — — — — — — —— — — — — —— —
1
I
I
I
I
I
I
I
I
I
I
I
I
\
I
I
I
I
1
I
I
I
I
I
\
provisional set
set
full
1.15
O
o
1.10
1.05
C22-C2B F =83.8
ksi
1-157-93/96
1.00
_i
I
I
I
1
I
I'll
I
_L
200
100
0
J
I
I
I
I
I
I
I
400
300
I
L
I
500
Strain rate, 1/s
15;56;34
•
Tuesday December
2004
14.
Figure C-5. Predictions of the provisional and full Cowper-Symonds model for perimeter
column spandrel steel C14-S, which is not in the provisional data set.
1
1
1
1
1
1
1
1
1
r
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
T
1
1
1
1
provisional set
"
o
set
full
O
o
1.2
/
'/
(
HH1 -WEB F =52.0
ksi
1-605-98/101
N
1
1
t
1
1
100
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
400
300
200
1
1
1
1
500
Strain rate, 1/s
1
5 :56
1
9
-
Tuesday December
1
4.
2004
Figure C-6. Predictions of the provisional and
column
268
steel
N99-C3T, which
full
is in
Cowper-Symonds model
for perimeter
the provisional data set.
NISTNCSTAR
1-3D,
WTC
Investigation
—
—
r
r
Provisional Analysis of High-Rate Data
Figure C-7. Predictions of the provisional and
column
N8-C1B, which
steel
I
1
1
I
[
I
I
I
- provisional set
-
set
full
1
[
— — — —— — — —
1
1
1
1
for perimeter
1
1
r-
——— —
1
1
1
.J-...,-.-P-.-I
1
o
O
o
Cowper-Symonds model
the provisional data set.
^
^^^^
1.8
full
is in
o
o
O
^
_
/
-
-/
-
C65-WEB
F =38.7
-904-86/89
1
5
1
1
1
1
100
0
1
1
1
1
1
1
200
1
1
1
1
1
1
1
300
1
1
400
1
1
1
ksi
1
1
1
1
1.
500
Strain rate, 1/s
15:56:14
-
Tuesday December
14,
2004
Figure C-8. Predictions of the provisional and full Cowper-Symonds model for perimeter
column wide-flange steel M26-C1B, which is in the provisional data set.
NISTNCSTAR
1-3D.
WTC Investigation
269
Appendix C
Figure C-9. Predictions of the provisional and full Cowper-Symonds model for core
column wide-flange steel HH-WEB, which is not in the provisional data set.
Figure C-10. Predictions of the provisional and full Cowper-Symonds model for core
column wide-flange steel C80, which is not in the provisional data set.
270
NISTNCSTAR
1-3D,
WTC Investigation
—
r
Provisional Analysis of High-Rate Data
-
———
1
I
'I
)
1
1
1
1
1
— nrnviQinnal
-
]
\
1
1
1
1
1
1
1
1
1
1
1
[ T'"l
cot
-
set
full
-
-
o
1.15
-
-
o
o
O
-
(0
-
o
-
-
-
-
-
O
"^"
^
-
/
C10-C1M1-IW F
{
= 112.5 ksr
1-451-85/88
\
1
1
1
1
1
100
0
1
1
1
1
1
200
1
1
1
1
1
1
1
300
1
1
400
1
I
1
500
Strain rate, 1/s
15:55:51
-
Tuesday December
14.
2004
Figure C-11. Predictions of the provisional and full Cowper-Symonds model for core
column steel C65-WEB, which is not in the provisional data set.
Figure C-12. Predictions of the provisional and full Cowper-Symonds model for core
column wide-flange steel C65-FL2, which is not in the provisional data set.
NISTNCSTAR
1-3D,
WTC
Investigation
111
Appendix
C
-
1
1
1
1
1
1
1
1
1.5
1
1
1
1
1
1
1
|--|-T--|
1
1
provisional set
set
1
c
1
1
1
o
full
:
-
1.4
-
—
—
o
-
o
-
-
-
-
-
_
-
Li
-
""^
^^O^
f
—
'
\J
"
—
/
i
"
M26-C1B F =62.5
[
ksi
1-130-90/93
\
1
1
1
1
)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
,
0
1
00
200
300
400
500
Strain rate, 1/s
15:54:54
-
Tuesday December
14,
2004
Figure C-13. Predictions of the provisional and full Cowper-Symonds model for core box
column steel B6152-2, which is not in the provisional data set.
272
NISTNCSTAR
1-3D,
WTC
Investigation
Appendix D
Deformation of Steels Used
WTC
in
7
INTRODUCTION
D.1
Unlike World Trade Center
(WTC)
1
and
WTC
no recovered
2,
steel in the
National Institute of
Standards and Technology (NIST) inventory can be unambiguously assigned to
properties
must be estimated completely from the
shapes conforming to two
rolled
columns were
specifications,
by
built
literature.
WTC 7. Therefore,
Building plans called for rolled column
ASTM International (ASTM) grades: A 36 and A 572
up with cover plates
thickness: (in note
/
<
that
were specified
2
A 36
A 588
A 572
A 588
in.
2in.
</<4m.
4in.
<r<6in.
/>6in.
About 26,000 tons of steel went
into
to several different
50.
The heaviest
ASTM
drawing S-17)
to
1
Grade
WTC 7 (Salvarinas
Grade 50
Grade 42
Grade 42
1986). Construction
documents collected during
the Investigation did not provide any information on steel suppliers, but a telephone interview with the
former project manager for the construction of WTC 7 yielded some information (Salvarinas 2003).
TradeARBED
Steel
the
41
supplied the
beams used
lb
jumbo columns
(now Corns) and Bethlehem
ft
These were rolled
for the floors.
(Salvarinas 2003).
ASTM
for the base of the building to
Steel supplied
They were supplied
to
A 36 and Grade 50
to a proprietary
steel.
Algoma shape
meet Canadian Standard
specifications. British
Algoma
CSA
Steel supplied
that
most of
was approximately
G40.21 ("General
Requirements for Rolled or Welded Structural Quality Steel/Structural Quality Steel") Grade 44W. This
was
a
Fy = 44
ksi steel,
The goal of this
report
where the
is
"W"
designates "weldable."
to estimate the relevant
mechanical properties of the
steels.
The
sections of this
appendix follows outline of main report on mechanical properties.
D.2
ELASTIC PROPERTIES
Chapter 2 of this report summarizes the values for the
Young's modulus or Poisson's
and 2-3 can be used
D.3
ratio differs
between
NIST found no
evidence that the
Chapter
2,
Eqs. 2-2
directly.
ROOM TEMPERATURE PROPERTIES
Because NIST recovered no
steel
The values of the parameters
that
from
WTC
make up
7,
it is
not possible to
make any
statements about
its
quality.
the expressions for the stress-strain behavior can be estimated
using the same methodology that was used for the
that
elastic properties.
structural steels, so the values in
WTC
1
and
WTC 2 steels.
This section summarizes
methodology.
NISTNCSTAR
1-3D.
WTC Investigation
273
Appendix
The
D
static yield strength, a^.^ for rolled
literature data
shapes was calculated using the expected historical value from
(Alpsten 1972):
D-1
also 3-3
where
F,
is
the specified
historically, the mill test
minimum
yield strength,
about 20 percent. (Alpsten 1972). The factor
to the value the static value.
specimen for the mill
The
=
factor
The
test report
factor
Afiange
came from
(T
Ap
.2
corrects for the fact that,
Adynamic
^
= -2.6
ksi corrects for the fact that, until recently, the
-3.6 ksi corrects the strength in the mill test report
the higher-strength web, but the flange represents the majority
of the load-carrying area. For core rolled plates the correction
where
1
report value of the yield strength of rolled shapes exceeds the specified value by
= 1.092 (Alpsten
ys
is
slightly different:
= k F + kdynamic
P y
alSO 3-4
1972).
The Voce hardening law can represent the increase
in stress, a,
with plastic
regime. In this hardening rule the flow stress, o, reaches a limiting stress,
strain,
s,,,
in the plastic
with an exponential decay:
,
D-3
also 3-5
This relation
is
available in most finite element packages.
Table D-1 summarizes the estimated
for the steels
from
WTC
static yield strengths
and the parameters of the Voce hardening law
7.
Table D-1. Parameters for calculating room-temperature stress-strain curves.
Steel Description
a,,
TS
ksi
ksi
ksi
Source of Plastic Portion
b
ksi
ksi
Fy = 36
ksi
WF shapes
36
36.8
61.3
0.190
30.337
24.467
67.973
S9-C1T-S1
36 ksi perimeter
Fy = 50
ksi
WF shapes
50
53.8
74.8
0.202
23.023
32.116
30.495
HH-FL-1
Fy = 36
ksi
cover plates
36
35.6
61.2
0.204
21.723
30.729
58.392
B6152-2-FL2
36 ksi box column
588 cover plates
2 in. < t < 4 in.
50
50.8
72.5
0.202
23.023
32.116
30.495
HH-FL-1
A
42
42.0
65.4
0.202
23.023
32.116
30.495
HH-FL-1
42
42.0
65.4
0.202
23.023
32.116
30.495
HH-FL-1
steel
50 ksi core wide-flange
t
<
2
in.
steels
A
4
572 cover plates
in.
A
<
t
<
6
in.
588 cover plates
t>6.0
Note:
in.
£n,ax is
the true strain at the tensile strength, TS.
The
labeled "source of plastic portion" identifies the specific
274
static yield strength, a,,„ is calculated
from Eq. D-2. The column
WTC specimen used to develop the voce hardening parameters
NISTNCSTAR
1-3D,
WTC
Investigation
Deformation of Steels Used
DA
in
WTC
7
HIGH STRAIN RATE PROPERTIES
WTC
Because, prior to collapse,
7 did not suffer any high strain rate events,
NIST made no
effort to
estimate high-strain-rate properties of the steel.
IMPACT PROPERTIES
D.5
WTC 7 did not suffer any high strain rate events, and because there were no
Because, prior to collapse,
reco\ ered steels to
test,
NIST made no
effort to estimate
impact properties of the
D.6
ELEVATED-TEMPERATURE PROPERTIES
D.6.1
Stress-Strain Curves
The elevated-temperature
The
can be estimated using the methodology of Section 6.4.3.
stress-strain cur\'es
stress-strain cur\'es for
0°C <
T< 700^
steel.
for all
WTC 7 steels
is
represented using a power-law work-
hardening model:
also 6-4
where a has
units
parameter Rts
is
of ksi, Thas units of °C and K(T) and n(T) are defined
the ratio of the
temperature tensile strength,
room temperature
TS,^f,
of the
steel
tensile strength
of the
in Eqs. 6-5
steel
and
,
D-5
=
also 6-7
TS^^f
corrects for
two
additional
room
used to develop the model expressed by Eq. D-4.
R
The parameter Rc
The
6-6.
of interest, TS, to the
phenomena. The temperature-dependent functions K(T) and
n(Tj (Eqs. 6-5 and 6-6) were originally developed using data that were not corrected to zero strain rate,
whereas the stress-strain cur\ es supplied to the Investigation teams for room temperature behavior are
corrected to the zero strain rate. Secondly, because the temperature dependence of
represented using the behavior of only two different steels,
it is
all
the possible steels
is
possible that the stress-strain behavior
room temperature could differ from that predicted by the akeady generated room
temperature stress-strain curve. The parameter Rc corrects the elevated-temperature stress-strain curve so
that the value of stress predicted by Eq. D-4 for T = 25 °C at e = 0.05 equals that of the room temperature
predicted by Eq.
at
model curve. In general,
Table
D-2 summarizes
the correction
is
small: 0.9
< Rc < 104.
the values of the parameters necessary to calculate high-temperature stress-strain
curves for the
WTC 7 steels.
NISTNCSTAR
1-3D.
WTC
Investigation
275
Appendix
D
Table D-2. Parameters for estimating elevatedtemperature stress-strain and creep curves.
TS
Steel Description
Parameter defined
ksi
Rc
Rts
ksi
6-7
in
6-19
Equation
Fy = 36 ksi
61.3
1.03
0.90
1.15
Fy
WF shapes
ksi WF shapes
36
=50
50
74.8
1.01
0.97
0.94
Fy
-36
ksi cover plates
36
63.4
1.07
0.86
1.11
50
72.5
0.98
0.97
0.97
42
65.4
0.88
0.94
1.08
42
65.4
0.88
0.94
1.08
/
<2
A
588 cover plates
2 in
A
4
<
/
<4
in.
572 cover plates
in
A
t
in.
<
t
<6
in.
588 cover plates
>
6.0 in
Note:
R(j
based on a reference tensile strength TSm49 = 70.65
ksi.
Creep Behavior
D.6.2
The creep behavior can be estimated using the methodology of Section 6.4.4. The creep strain as a
function of time is represented by Eq. 6-10 using the parameters of Eqs. 6-11, 6-12, and 6-13.
D-6
s^.=At'a'
also 6-10
The applied
reference
stress
should be scaled by R^, ratio of the tensile strength of the steel in question to that of the
AS A149
steel (r5'Ai49
= 70.6
ksi) using Eq. 6-19.
c
=R(J^,a =
TS A
1
D-7
49
(J
(7
also 6-19
TS..
Table
D-2 summarizes
for the
the values of the parameters necessary to calculate high-temperature creep curves
WTC 7 steels.
D.7
REFERENCES
D.7.1
References Available from Publicly Available Sources
Salvarinas, John
J.
1986. Seven
World Trade Center,
New York,
Fabrication and Construction Aspects.
Canadian Structural Engineering Conference-] 986 Proceedings. Canadian
Willowdale, Ontario, Canada.
276
ISBN 0-88811-062-6
pages
1
1-1 to
1
Structural Steel Council,
1-44.
NISTNCSTAR
1-3D,
WTC Investigation
Deformation of Steels Used
D.7.2
in
WTC
7
References Available from Nonpublic Sources
Salvarinas. John
J.
2003. Telephone interview with
for Frankel Steel during the erection
NIST NCSTAR
1-3D,
WTC
Investigation
of
WTC
7.
WilHam Luecke.
May
Salvarinas
was
the project
manager
16.
277
Appendix
D
This page intentionally
278
left
blank.
NISTNCSTAR
1-3D,
WTC
Investigation
Appendix E
Specimen Geometry Effects on High-Rate Tensile Properties
introduction
E.1
The
test
plan utilized several different specimen sizes and geometries to evaluate the strain rate sensitivity
of the yield and tensile strength and the
an attempt
to
minimize the ringing
from ordinary
test
total elongation.
Two
in the load signal that
comphcated
specimens for assessing the quality of the
temperature stress-strain behavior already existed.
convenient to combine these data
sets.
To
types of high-rate specimens were used in
steel
and
the data analysis. In addition, data
for characterizing
evaluate the strain rate sensitivity,
room-
was
it
This appendix describes the limitations of combining data
developed from different specimen geometries and sizes and summarizes the calculations of the
dependence of the
The
total
analysis reaches
two conclusions. Results from specimens with
to evaluate the strain rate sensitivity
the total elongation, the specimens
different
may be
different size, but they
a first approximation, the details of the
must be geometrically
of tension
tests to
determine material properties for use in
constitutive
models would be questionable.
E.1.2
Estimating Specimen-Geometry Effects on Ductility
total
self-similar.
specimen shape should not affect the measured yield and
tensile strength. If they did, the applicability
The
geometry can be combined
of yield and tensile strength. To evaluate the strain rate sensitivity of
Estimating Specimen-Size Effects on Yield and Tensile Strength
E.1.1
To
strain rate
elongation to failure.
elongation to failure of a specimen depends on the gauge length. In specimens with small gauge
lengths, the
neck occupies a greater fraction of the
total elongation,
which leads
to greater
measured
total
elongations. In general, geometrically similar specimens develop similar total elongations. This relation
called Barba's law (Metals
A(,
and gauge length
Handbook, 1985), which
relates the total elongation, £/,, to the
specimen area
Lo:
El,-/3-^ + e^
The parameter p
is
a proportionality coefficient, and e„
is
E-1
the engineering strain at the point of necking.
Generally, specimens with different geometries will produce similar total elongations
Eq. E-1
is
specimens
if the first
held constant. This relation underlies the choice of the geometries of the various
in
ASTM
NISTNCSTAR
1-3D,
is
E
8,
WTC
where the standard- and sub-size specimens have
Investigation
^"'^/Z-o
=
term
in
round
0.177.
279
Appendix
E
PROCEDURE
E.2
In the course of the
different
World Trade Center (WTC)
specimen geometries for
tests at quasi-static rates to satisfy the constraints
material and the test equipment available.
specimen geometries, were
employed
Investigation, investigators
The
quasi-static strain rates,
range 6x10""
in the
used the ordinary tensile specimens followed E
s"'
8,
<
ds/dt
<
1.2x10"^
which allows
of
of the starting
which were not
s"'.
a variety
Some of the
identical
between
quasi-static tests that
for an increased extension rate after the
yield strength has been determined. For these specimens, the strain rate plotted
is
the rate appropriate for
the tensile strength determination.
Table E-1 summarizes the different specimen geometries used for determining strain rate
Note
that the
/
=
0.25
Flat 2"
in.
A^'~/Lo ratio, but their
sensitivities.
specimen and the Flat-HSR-l'TK specimen not only have the same
gauge lengths have the same width
They
to thickness ratio.
are geometrically self-
similar.
Table E-1. Specimen geometries used for establishing strain rate sensitivities.
Specimen
Specimen
Description
Designation
High
rate
High
rate
3 in.
1
in.
TK
Flat-HSR-1" IN
Flat-HSR-1"
round for
Rd
A
H'
in.
in.
in.
in.
in.'
0.125
0.25
1.25
1
0.031
0.177
0.0625
0.25
1.25
1
0.016
0.125
n/a
1.5
1
0.049
0.220
D
1"
=
0.25 in
elongation
measurement
8
in.
long
flat'
Flat 2"
0.250
0.50
2.75
2
0.125
0.177
8
in.
long
flat''
Flat 2"
0.5625
0.50
2.75
2
0.281
0.260
4
in.
long
flat
Flat-C 1"
0.250
0.25
1.1
1
0.0625
0.250
in. specimens was generally 0.25 in.
Specimen M26-C1B1-RF came from a plate with / = 9/16 in.
Key: n/a, not applicable; A, area; L„, length of uniform cross section;
specimen gauge section width.
a.
Material thickness for Flat 2
b.
Lo, elongation
gauge length;
specimen thickness; w,
/,
RESULTS
E.3
Figures E-1 through E-7
show
In general, the agreement
between the
the quasi-static stress-strain curves for the different specimen geometries.
strength and tensile strengths. Table
different
E-2 summarizes
should be taken from the table, because
failure.
Although the yield and
are small
enough so
specimen geometries
the data
in soine cases the
good
from those
for
1
percent offset yield
figures. Total elongations
operator removed
tensile strengths are not identical
that the differences can
is
the extensometer prior to
between the specimens, the differences
be ignored and the data combined to evaluate the strain rate
sensitivity.
The
total elongation data generally
with larger
^
'
^/Lfi is
follow Barba's law. Fig. E-9. The total elongation, El„ in specimens
correspondingly longer
comparison between the
large, Flat-2"
280
which
is
of the six cases. Specimen
specimen and the 1/4-size Flat-HSR
ratio is identical for both test specimens.
indistinguishable,
in five
The
total elongations for
MlOB
provides a direct
TTK specimen. The A^ '/Lq
each specimen are
statistically
consistent with the expected Barba's law behavior.
NISTNCSTAR
1-3D,
WTC Investigation
Specimen Geometry
hsr
hsr
Isr
Isr
Effects
on High-Rate Tensile Properties
M10B-C3B1-L4
M
iOB-C3B1-L5
M10B-C3B1-L1
M10B-C3B1-L2
20
Flat 2
in.
= 0.125
in.
Isr
hsr
0
_l
1
J
L.
1
0.0
I
I
0.2
0.1
L.
J
t
I
I
I
1
0.4
0.3
Engineering Strain
12:57:04
•
Tuesday December 21 2004
.
Figure E-1. Comparison of quasi-static stress strain curves obtained with high-rate style
(HSR) and ordinary specimens for specimen M10B-C3B1-RF, Fy = 100 ksi.
100
1
1
1
1
r
C/D
O)
c
I
c
"d>
c
hsr
Isr
40
Isr
N99-C3
N99-C3
N99-C3
LU
20
2
in.
= 0.0625
in.
Isr Flat
hsr
J
I
L.
0.0
_l
0.1
I
I
t
L.
0.2
0.3
0.4
Engineering Strain
12:57:01
-
Tuesday December 21. 2004
Figure E-2. Comparison of quasi-static stress strain curves obtained with high-rate style
(HSR) and ordinary specimens for specimen N99-C3M1-RF, Fy = 65 ksi.
NISTNCSTAR
1-3D,
WTC Investigation
281
Appendix
E
100
\
I
I
n
CO
CO
C/)
Q)
CO
CD
_c
0)
CD
N8-C1B1-L2
N8-C1B1-LF
hsr
40
c
'O)
c
LU
20
ordinary specimen
hsr
_^
0.0
0.1
0.2
,
\
t
LF!
is
2
in.
= 0.0625
in.
Isr Flat
,
,
I
I
0.3
0.4
Engineering Strain
12;57:02
•
Tuesday December 21 2004
,
Figure E-3. Comparison of quasi-static stress strain curves obtained with high-rate style
(HSR) and ordinary specimens for specimen N8-C1B1-RF, Fy = 60 ksi.
Figure E-4. Comparison of quasi-static stress strain curves obtained with high-rate style
(HSR) and ordinary specimens for specimen M26-C1B1, Fy = 50 ksi.
282
NISTNCSTAR
1-3D,
WTC Investigation
Specimen Geometry
Effects on High-Rate Tensile Properties
Figure E-5. Comparison of quasi-static stress strain curves obtained with high-rate style
(HSR) and ordinary specimens for specimens from the flange of WF core shape C65-FL2,
Fy = 36 ksi.
100
80
CO
CO
o
CO
D)
CD
0}
c
'U)
c
LU
20 Round
Isr
hsr
0
.J
I
J
L.
0.0
0.1
I
J
L.
0.2
0.3
t
<
1
= 0.125
<
<
in.
in.
1
0.4
Engineering Strain
12:57:04
-
Tuesday December 21 2004
,
Figure E-6. Comparison of quasi-static stress strain curves obtained with high-rate style
(HSR) and ordinary specimens for specimens from the web of WF core shape C65-WEB,
Fy = 36 ksi.
NISTNCSTAR
1-3D,
WTC
Investigation
283
Appendix
E
Figure E-7. Comparison of quasi-static stress strain curves obtained with high-rate style
(HSR) and ordinary specimens for specimens from core box column B6152-2 Fy = 36 ksi.
284
NISTNCSTAR
1-3D,
WTC Investigation
—
J
r
^
J
3
"
1
1
K
"
Specimen Geometry
Table E-2.
Summary
1
on High-Rate Tensile Properties
Effects
data for tests using different specimen geometries at
quasi-static rates.
Strain
Strain
Rate for 1% Rate for
Fy
offset
TS
TS
El,
ksi
L-ci
/o
in.
O J. J
7A A
30-U
-)
oD.
1 J.U
s A
3
-)
1/s
1/s
7
1
1
O.VOII-J
J
A
0
n—
-U
OAF
f,
If.
.u
/
c.
—
D.UOt-j
I
.
1
1
I
1
.
I
fs.3
11^
^
jt-j
A7F-^
I
J
\
QF
711 —
'i'^F
fi
1
ut
ri'^t
1-41
-U
/
1
1 7F
1.1/
t-J1
"^AF 1
1
5.
/
J t-j
1
1
1
1
1
1
Q
-I-.
J
A
.0
/h-4
.U
Jt-J
70 A
-)
1
7
I
p
Qzl
0
TA A
->
1
n7P
1
1
Q
t!
jy.o
4U.fS
J V.o
17/.D
A
J
17
Q
1o o
JO.O
17 C
38
3
1
77 A
7
1
p I
n-j
I
1
P
C-J7
CCP
1
A
HP
1
JC-
Js.j
Jh-j
n7F
1
.U
I
I
/I
1
/
1
fc-4
7P
1
7 AP 7
17P
A
4.
1
A C
"7
_3 .4
_U.4
1
7A
1
7
1
A
">
1
I ">
_4.4
I
I
1
1
1
7
'I
1
1
V.J
fl A
iv.v
c c
O.J
A
A
o.U
0
->
77/.U
A
-
I
_ J.U
1
1
1
1
A
An
du
A
AA
00.
^A A
JO.O
A<
DJ. 7/
7<
A
J.U
3
OJ.4
lA A
3d.U
03. J
4
64.0
60.9
3
1
1
K.
Ivl26-C
25
65
Hat
HSR
1
1
N
N99-C3]vl -Rh-L6
1-148-99/102
N9y-L3
-Rr-Ll
1
N99-L J
-RF-L2
1
1
1
A
in
U.I//
65
Flat 2"
A
T Tl
U. / /
65
Flat 2"
77
85
Flat
HbK
HSR
1
77
85
Flat
77
85
Flat 2"
AT
6z
0.25
0.
1
77
85
Flat 2"
A "7/
0.25
0.250
90
Flat-C
1
0.25
0.250
1
0.25
i
^1
j3
A AAT C
1
1
1
K
B -L
1
1
M
M
i
I
1
1
DC A
OB-L 3B -RF-L4
1
IJH
T A AA A 1
-130-90/93
I
\
1
1
/
AO AA AT
l-l4;s-99/IUz
1
\/Il/lD
1
1
]
\
/ 1
AA/IAT
-148-99/1
02
1/10
">
")AA 01/0A
2-/06-83/86
OB-C 3B -RF -L5
2-206-83/86
MI0B-C3B1-Rr-Ll
2-206-83/86
M
2-206-83/86
C
1
0B-C3B 1-RF-L2
1
1
0-C M-FL-l
1
-1
-45 -85/88
1
1
90
Flat-C
L lO-C Ivl-rL-1-2
1
/let OC/OO
-45
-85/88
0,
77
nA
1
yu
Flat 2"
L U-L D-rL-l -LI
1
1
/1C1 OC/OO
-43
-83/50
0.
1
77
90
Flat 2"
ClO-C lB-hL-l-L2
A c
OC/OO
1-451-85/88
A
"K
U. Zj
1
1
1A
JO
r lat
HSK
Flat
HSR
1
5
0.0625
0.
125
36
AO
A
C
U.J
A
on
1
-)
A"?
A
/
1
1
t
1
-)
0
1 A A A /A 1
1-130-90/93
HbK
1
1
1
A A /A1
Hat
1
66
c\
50
1
0.
6
")
30-90/93
Flat 2"
0.
1
1
-1
1
50
1
0.25
C
A
36
Rd
36
Rd
1
1
1
1
1
IN
DO j2-i-rL-2-L4
1
1
N
B61 j2-2-rL-2-Lj
i
2"
bol jz-i-rL-2-L
2"
DAIO
TCI Tl")
Dol j2-z-rL-2-L2
1
1
y
1
I
/O A
-jU4-33/3d
1
-504-33/30
C A/I
1
C A/1 11 /I ^
-jU4-33/3o
1
1
"3
I
C A/1 1 T /I A
-jU4-33/3d
1
1
AA
66
0.25
0.222
36
Rd
1
CDJ-2-WhD-Ll
1
1
4_.U
0.25
0.222
36
Rd
Loj-2- whb-L2
1
1
63
1
1
AA A O £ /O A
-yu4-oo/oy
TIC
.J
]
55
0.0625
25
36
Flat
L65-WtB-L4
1
1
AA A O /O A
-904-86/89
0.222
36
Rd
C65-H.25-LI
1
v. -21
36
Rd
1
''"'2
Jv
Rd
1
1
.U
1
/
AC
A
OJ.U
4_.U
1
0
4 37E-4
70 0
43 0
]
62
4.3
M26-C Bl -RF-L2
IBI-RF-Ll
1
1
A
11
U.I//
55
66.6
/
0,265
A
1C
U. 25
/I
7 "TC
c-^A
/I
lvI26-C
1
A 11X1
1-4A
4.3
Flat 2"
0.
60
O.U
J 3.
T^
1
Location
ff
50
1
A
A
0
A
0
0^
sample
Specimen type
0.265
0.
An
AA
A
C
IZI
1
J
I
04. J
1
1
0.5625
U.
c
-)
fy
ksi
1
I
1
7A A
J
1
O. J
I
I
i
1
I
"> t
I
I
I
U.J
n^p
1
1
1
/I
1
1
7
n "7/
jy.
/
1
1 A
J.D
1
O.
Q
J.J
jy. J
1
QC 0
p
L-J
5
0
/
in.
n CATC
I
1
1
I
.o
/o
61
1
1
1
OQ Q
1
77 A
—7
I
">
7
J
"''11
1
1
p
07F-A
1
-)
^.J"
1
1
1
7
1
7";
1
U
El-J
1
1
1
<,
D.UDt-J
f\
70 R
1
1
7
1./
1
70
07E-4
I
I
1
ROA
43.0
1
63
0.25
A
A
1
I
0
C
""5
0.
0
1
HSR
1
IN
I
(-
65-H.25-L2
PAS H
I
A A A OA /OA
-yu4-co/oy
1
1
1
1
1
AAyl OA /OA
-yu4-86/8y
AAyl OA /OA
-yU4-a6/8y
QAA SA/8Q
8.75E-5
47.0
4.37E-4
68.0
38.0
1
63
0.25
0.222
36
Rd
1"
C65-H.5-L1
1-904-86/89
8.75E-5
42.5
4.37E^
66.0
42.0
1
62
0.25
0.222
36
Rd
1"
C65-H.5-L2
1-904-86/89
C65-H.5-L3
1-904-86/89
C65-H.5-L4
1-904-86/89
TN C65-FL2-L4
1-904-86/89
8.75E-5
37.1
4.37E-4
64.0
43.0
1
62
0.25
0.222
36
Rd
1"
2.19E-3
38.7
2.19E-3
65.6
42.0
1
62
0.25
0 222
36
Rd
1"
1.07E-4
40.1
1
54
65.3
39.2
0.0625
0.125
36
Flat
HSR
DISCUSSION OF TOTAL ELONGATION BEHAVIOR
E.4
The
I.07E-4
1"
excellent correspondence between the different-sized specimens of
possible, the different sized specimens with identical
/i' '^/Lo
ratios
M
1
OB
indicates that,
where
can be combined to gain insight into
the behavior of the total elongation with strain rate.
E.4.1
Limitations
Although the specimens are geometrically
in the
similar, the analysis
specimen as the work of deformation
NISTNCSTAR
1-3D,
WTC Investigation
is
does not account for the temperature
transformed into heat. High rate
tests are generally
rise
assumed
285
Appendix
to
E
be adiabatic. Quasi-static
temperature rise will be
less.
tests
have additional time for the specimen to cool during deformation, so the
For adiabatic conditions the temperature
of (?/. For
C;,
= 415
steel,
assume
J/kg/K, and p
that the failure strain
is
\s(e)de
p
'
C,,,
AT,
1
Ar =
for a material with density, p, heat capacity
rise,
e
which deforms
=
is
E-2
0.3, the
at
engineering stress s(e) to a final strain
flow stress constant s = 830
MPa (120 ksi),
= 7860 kg/ml Then
—^=76.5K
S6
AT =
which represents an upper
limit.
The temperature
rise
E-3
of lower-strength specimens will be
correspondingly smaller. Because the temperature rise of the quasi-static specimen
the effect of the temperature rise on the ductihty of the high-rate specimens
E.5
It is
not zero, however,
probably not significant.
CONCLUSIONS
possible to combine ordinary and high-rate style specimens for determining the yield and tensile
behavior as a function of strain
rate.
For determining the
total
elongation behavior as a function of strain
because of the Barba's law behavior, specimens can be mixed only
are identical and their ^' "/Lo ratios are identical.
rate,
E.6
is
is
if their
width-to-thickness ratios
REFERENCES
Metals Handbook. 1985. Metals Handbook Ninth Edition. Volume 8 Mechanical Testing. American
Society for Metals. Metals Park,
286
OH. USA.
p. 26.
NISTNCSTAR
1-3D,
WTC
Investigation
»
Specimen Geometry
Effects
on High-Rate Tensile Properties
O
:
D
I-
A
F^,=85 ksi
0
F,.=90 ksi e. =(66.2613.42) ClO-CilVi-FL
Fy=36
r,=36
F„=36
150
CO
N99-C3M1-Rr
,.=65 ksi e. =1-85. 91±1 5.05}
ksi
M10B-C3B1-RF
e,=(2.59±8.43) B6152-2-FL
ksi e, =(29. 50x4.48)
ksi
C6S-web
e,={11.98±22.16) c65-flange
-0
C
0)
^
100
c
50
0.10
15:04:56
-
0.20
0.15
0.25
Thursday December 16 2004
T
17.
^50
F =50
ksi
F =65
ksi
F,.=85 ksi
F'=90
Fy=36
M10B-C3B1-RF
C10-C1M-FL
ksi
e-,=f41.67±6.30i
ksi
e,=(-15.53±12.97) B6152-2-FL
e,=(-i .04114.34) C65-web
ei=(0. 59138. 60) c65-fiange
F.,=36 ksi
f'=36
ei=(-92.19±14.70) M26-C1B1-RF
e-=(-74.32±51.83) N99-C3tVi1-RF
ksi
c
0}
i—
•—
p
100
0)
CD
(0
50
J
I
L
I
0.10
15:04:56
Figure E-8. Yield
-
Thursday December
(1
1-3D,
WTC
16.
I
I
L
0.20
0.25
2004
percent offset) and tensile strength as a function of specimen
geometry
NISTNCSTAR
_l
0.15
for
Investigation
specimens tested
at quasistatic rates.
287
Appendix
E
60
o
F =50
ksi
e,=(145.93±9.80)
M26-C1B1-RF
N99-C3Mi-RF
l\=(^5 ksi c,=(104.:3±!6.72)
A
50
M10B-C3Bi-RF
F;=85
ksi
F;=90
F,=36
F;=36
ksi e,=(
F =36
95.63127.32) Cl(t-C1M-FL
ksi e,=(5.69±17.85)
B6152-2-FL
e.=(l03.52±8.97) C65-wcb
ksi e,=(27.5 1+19.62) c65-flange
k.si
*
40
•4—*
O
H
20
10
I
—— —
'
I
l5:l).S:(lf) -
Tllursday
Dcccmhcr
0.15
I
I
L.
0.20
0.25
16, 20(14
Figure E-9. Total elongation,
288
_1
>
0.10
Elt, as a function of specimen geometry for specimens
tested at quasistatic rates.
NISTNCSTAR
1-3D.
WTC
Investigation
Download