Report on Methods
for Estimating In-Place
Concrete Strength
Reported by ACI Committee 228
I
0:::
•
co
N
N
u
<(
�
�
American Concrete Institute
Always advancmg
American Concrete Institute
Always advancing
irst Printing
January 2019
ISBN: 978-1-64195-050-3
Report on Methods for Estimating In-Place Concrete Strength
Copyright by the American Concrete Institute, Farmington Hills, MI. All rights reserved. This material
may not be reproduced or copied, in whole or part, in any printed, mechanical, electronic, film, or other
distribution and storage media, without the written consent of ACI.
The technical committees responsible for ACI committee reports and standards strive to avoid
ambiguities, omissions, and errors in these documents. ln spite of these efforts, the users of ACI
documents occasionally find information or requirements that may be subject to more than one
interpretation or may be incomplete or incorrect. Users who have suggestions for the improvement of
ACI documents are requested to contact ACI via the errata website at http://concrete.org/Publications/
DocumentErrata.aspx. Proper use of this document includes periodically checking for errata for the most
up-to-date revisions.
ACI committee documents are intended for the use of individuals who are competent to evaluate the
significance and limitations of its content and recommendations and who will accept responsibility for
the application of the material it contains. Individuals who use this publication in any way assume all
risk and accept total responsibility for the application and use of this information.
All information in this publication is provided "as is" without warranty of any kind, either express or
implied, including but not limited to, the implied warranties of merchantability, fitness for a particular
purpose or non-infringement.
ACI and its members disclaim liability for damages of any kind, including any special, indirect, incidental,
or consequential damages, including without limitation, lost revenues or lost profits, which may result
from the use of this publication.
It is the responsibility of the user of this document to establish health and safety practices appropriate
to the specific circumstances involved with its use. ACI does not make any representations with regard
to health and safety issues and the use of this document. The user must determine the applicability of
all regulatory limitations before applying the document and must comply with all applicable laws and
regulations, including but not limited to, United States Occupational Safety and Health Administration
(OSHA) health and safety standards.
Participation by governmental representatives in the work of the American Concrete Institute and in
the development of Institute standards does not constitute governmental endorsement of ACI or the
standards that it develops.
Order information: ACI documents are available in print, by download, through electronic subscription,
or reprint and may be obtained by contacting ACI.
Most ACI standards and committee reports are gathered together in the annually revised the ACI
Collection of Concrete Codes, Specifications, and Practices.
American Concrete Institute
38800 Country Club Drive
Farmington Hills, MI 48331
Phone: +1.248.848.3700
+1.248.848.3701
Fax:
www.concrete.org
ACI228.1R-19
Report on Methods for Estimating In-Place Concrete
Strength
Reported by ACI Committee 228
Andrew J. Boyd. Chair
B mard H. Hertlein.Secretary
Todd Allen
Boris Dragunsky
Muhanuned P. A. Ba heer
Christopher C. Ferraro
Michael C. Forde
RobertS. Jenkins
Liying Jiang
Eric R. Giannini
Keith E. Kesner
W i l liam Ciggelakis
Aldo De La Haza
Kerry
. Hall
Hai
Julie Ann HaneU
Ethan C. Dodge
Larry D. Olson
Michael W. Hoag
Mosta(a Mohamed Gad Alia
Michael D. Brown
icholas J. Carino
Frederick D. Heidbrink
Stephen Pes iki
JohnS. Popovics
athaniel Steven Rende
Patrice Rivard
. Lew
Patll L.
Malcolm K. Lim
iwek
Patrick J. E. Sulli an
Consulting Members
John H. Bungey
Honggang
1-lermenegildo
•
Kenneth M. Lozen
Gerardo G. Clemena
ao
eilA.
aratin
AI Gh rhanpoor
Clau� Germann Petersen
George V. Teodoru
Alexander M. Le hchinsky
estor E. Chonillo
Herbert Wiggenhauser
V. M. Malhotra
umming
ubcommittee Member
hris M. McDenn
u
This report provides methods for esh"maling the in-place strength of
concrete in new and existing com/ruction. These methods include:
rebound numbet; penetration resistance. pullout. pull-off, ultra­
Keywords: coefficient of variation; compressive strength; construction
safety; in-place tc ts: nondcstructi c test : sampling; statistical analysis.
CONTENTS
sonic pulse velocity, maturity, and cast-in-place cylinders. The
principle. inherent limitations, and repeatability of each method
are reviewed. Pmcedures are presented for developing the relation­
ship needed to estimate compressive strength fi"Om in-place results.
Factors to consider in planning in-place tests are discussed, and
statistical techniques to inte1pret lest results are presented. The
use of in-place tests for acceptance of concrete is inft"Oduced.
The Appendix A pmvides information on the number of strength
levels that should be used to develop the strength relationship and
explains a regression analysis procedure that accoums for ermr in
both dependent and independent variables.
C HAPTER 1 -INTRODUCTION, p. 2
I . I -Scope,p. 2
1 .2- eed for in-place te ts during con tmction, p. 2
1 .3-fnftuence of A T 3 1 8 , p. 3
1 .4-Rccommendations in other ACT documents, p. 3
1 .5-Exi ting constm rion, p. 4
1 .6-Report obj ctivc, p. 4
C HAPTER 2-NOTATION AND DEFINITIONS, p. 4
2. 1 - 1otation, p. 4
2.2-Definitions, p. 5
ACJ Committee Reports, Guides, and
ommentaries are
intended for guidance in planning, designing, executing, and
inspecting construction. This document i intended for the use
of individual who arc competent to evaluate the significance
and limitations of its content and recommendations and who
will accept responsibility for the application of the material it
contains. TI1e American
oncrete In titute disclaims any and
C HAPTER 3-REVIEW OF M ETHODS, p. 5
. 1 -Tntroduction, p. 5
3 .2-Rebound number {A TM 805/ 805M), p. 6
3 .3-Pcnctration resistance (A TM 803/C803 M ). p. 7
3.
Pullout te r (A TM C900), p. 8
all respon ibility for the stated principles. The Institute shall
not be liable for any lo s or damage arising therefrom.
Reference to this docmnent
documents. If item
hall not be made in contract
found in this document are desired by
the Architect/Engineer to be a part of the contract documents,
they shall be restated in mandatory language for incorporation
by the Architect/Engineer.
ACI
2019.
228.1 R-19 supscrscdes ACI 228.1 R-03 and was adopted and published January
opynght
2019. American Concrete Institute.
All rights reserved including rights of reproduction and use in any form or by
""Y means, including the making of copies by any photo process, or by eleclronic
or mechanical dcvu�:c, prinled. wnttrn. or oral, or recording for sound or visual
reproduction or for use in any knowledge or retrlc al
sys tem
pennis ion in writing is obtained from the copyright proprietors.
or dev1ce, unlc s
2
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228. 1 R- 1 9)
3 .5-Pull-off test (A TM C 1 583/C1583M ), p. I I
TM C597), p. 12
3 .6 U ltrasonic pulse velocity (
3 . 7 Maturity method (
TM C 1 074), p. 1 4
3.
TM 873/ 73M ), p. 1 6
ast-in-place cylinders (
trength limitation , p. 16
3.
3 .10
ombined method . p. 16
3 .11
ummary, p. 17
CHAPTER 4-STATISTICAL CHARACTER ISTICS
OF TEST RESULTS, p. 1 7
4.1- eed for tatistical analysis, p. 1 7
4.2-Repeatability of test result , p. 1 8
CHAPTER 5-DEVELOPM ENT OF STRENGTH
RELATIONSHIP, p. 23
5 .1
5.2
5 .3
General, p. 23
ew construction, p. 24
Existing con truction, p. 27
CHAPTER 6-IM PLEM ENTATION OF IN-PLACE
TESTING, p. 28
6. 1- ew construction. p. 28
6.2-Existing con truction, p. 31
CHAPTER ? -INTERPRETING AND REPORTING
RESULTS, p. 32
7.1
7.2
7.3
General, p. 2
tatistical method , p. 33
Reporting re ults, p. 36
CHAPTER S -IN- PLACE TESTS FOR
ACCEPTANCE OF CONCRETE IN NEW
CONSTRUCTION , p. 38
8.1- eneral, p. 38
8.2-Acceptance criteria, p. 38
8.3 Early-age testing p. 3
CHAPTER 9-REFERENCES, p. 39
Authored documents, p. 40
APPENDIX A , p. 44
A. l M i nimum nw11ber of strength level , p. 44
A.2 Regression analysis with X-error (Mandel 1984), p. 44
A.3- tandard deviation of estimated Y-value ( tone and
Reeve 19 6), p. 46
A.
Example, p. 46
CHAPTER 1 -I NTRODUCTION
1.1 -Scope
In-place te ts are performed typically on concrete within a
tructure, in contrast to te t performed on molded specimen
made from the concrete to be u ed in the tructure. Hi Iori­
cally, they ha e been cal led nonde tructive tests becau e
some of the early tests, such as rebound number and uln·a­
sonic pul e elocity, were noninvasive and did not damage
the concrete. Over the years, howe er, new method have
developed that re ult in superficial local damage. There-
fore, the tem1inology "in-place tests" is used as a general
name for these test methods, which includes those that do
not damage the concrete and those that result in some nearurface damage. In thi report, the principal application of
in-place tests is to estimate the compressive trength of the
concrete. The pull-off test can be used to e tim ate the tensile
trength of concrete or evaluate bond trength between
layers. The ignjficant characteristic of most of the e tests i
that they do not drrectly measure the compre ive trength
of the concrete in a structure. In tead, they mea ure ome
other property that can be correlated to com pres ive trength
(Popovics 1998). The strength is then estimated from a
previously establi bed relationship between the measured
property and concrete strength. The uncertainty of the esti­
mated compre sive trength depends on the ariability of
in-place test results and the uncertainty of the relationship
between these two parameters. The e ource of uncer­
tainty are di cussed in this report. An alternative approach
for correlation between te ts re ults and concrete trength i
pre ented in E I 3 79 I (2007) and B 60 9 (20 I 0).
In-place test can be used to e timate concrete trengtb
during construction so that operations requiring a specific
strength can be performed safely or curing procedures termj­
nated. They can also be used to e timate concrete trength
during the evaluation of existing structures. These two
applications require l ightly different approaches, so parts of
this report are eparated into section dealing with new and
exi ring con truction.
A variety of technique are a ailable for estimating the
in-place trength of concrete ( Malhotra 1976; Bungey et
al. 2006; Malhotra and Carino 2004). o attempt i made
to review all method in thi report· only tho e method
that have been tandardized by A TM I nternational are
discussed. Examples of methods not co ered include internal
fracture test (Chabowski and B1yden- mith 19 0; Domone
and Castro 1987) and torque te t ( toll 1985 ).
1.2-Need for in-place tests during construction
In North American practice, the mo t widely used test for
concrete is the compressive strength te t of standard cylin­
der ( STM C39/39M ). This te t procedure i relatively
ea y to p rform in t rm of ampl ing, pecimen preparation
( STM C31/C31 M), and trength measurement. hen prop­
erly performed this test has low single-operator variation
and low interlaboratory variation and therefore, the method
lend itself to use as a standardized te ting procedure. The
compressive strength so obtained is u ed to verify that the
specified strength (f;') used to calculate the nominal strengths
of tructural member has been achieved. Therefore the
compres ive strength of tandard cylinders is an e sential
parameter in design code and project specifications.
hen carried out according to tandard procedure ,
howe er, the re ult of the cylinder compres ion te t repre­
sent the poten6al strength of the concrete as delivered to a
site. The test is u ed mainly as a basis for quality assurance
of the concrete to ensure that contract requirement are met.
It is not intended for determining the in-place strength of
the concrete because it makes no allowance for the effects
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228. 1 R-19)
of placing, consolidation or curing. It is unusual for the
concrete in a structure to have the same properties as a
standard-cured cylinder at the arne test age. lso standard­
cured cylinders are usually tested for acceptance purpo es at
an age of28 days· therefore, the result of these te ts cannot
be used to determine whether adequate trength exists at
earl ier ages for safe removal of formwork or the application
of po !-tensioning. The concrete in some part of a truc­
ture, uch a colwnn may develop trength equal to the
tandard 28-day cylinder trength by the time it i ubjected
to design loads. Concrete in flexural member , especially
pretensioned members, can be required to support a large
percentage of the design load before an age of 28 days. For
these reason , in-place tests are used to estimate the concrete
trength at critical location in a structure and at times when
crucial construction operations are scheduled.
Traditional ly, a mea ure of the trength of the concrete in
the structure ha been obtained by using field-cured cyl inders
prepared and cured in accordance with A TM C3 1 /C3 1 M .
The e cyl inder are cured on or in the tru ture under, a
nearly a possible, the arne condition as the concrete in the
tructure. Measured strengths of field-cured cylinders may
be significantly d ifferent from in-place strengths because it
i difficult, and often impossible, to ha e identical bleeding
con olidation, and curing condition for concrete in cylin­
der and concrete in structures ( outsos et al. 2000). Field­
cured pecimens should be handled with care and tored
properly to avoid misleading test result .
Con truction schedules often require that operation uch
a form removal, post-tensioning, termination of curing, and
removal of re hore b carried out a early a po ible to
keep the project on chedule. To enable the e operation to
proceed afely at the earliest pos ible time requires the use
of reliable in-place tests to estimate the in-place strength.
The need for uch strength information is emphasized by
everal con truction fai lure that po ibly could have been
prevented had in-place testing been u ed ( Lew 1 9 0; 'arino
et al. 1 9 a). In-place te ting not only increases safety but
can re ult in substantial co t saving by permitting acceler­
ated construction schedules ( Bickley 1 982a).
1.3-lnfluence of ACI 3 1 8
Before 1 9 3 , ACI 3 1 required testing o f field-cured cylin­
ders to demonstrate the adequacy of concrete strength before
removal of formwork or re horing. l n 1 983, ACI 3 1 8 first
allowed the use of other procedures instead of tests for field­
cured cylinders, if approved by the building official (
I
ommittee 3 1 8 1 983). The design professional, when requested
by the building official however, wa required to approve the
altemative procedure before its u e. Since 1 9 3, A I 3 1 has
permitted the u e of in-place testing as an alternative to te ting
field-cw-ed cylinders if approved by the licen ed de ign profe ional and, if reque ted, approved by the building official. The
commentaiy to ACI 3 1 - 1 4 ( ection R26. 1 1 .2. 1 (e)) l i ts fow­
procedures, which are covered in this report, that may be used
pro ided there are sufficient correlation data.
Most design provision in
I 3 1 are based on the
compressive strength of standard cylinders. Thu to evaluate
3
tructural capacity under construction loading 1t IS nece ary to have an estimate of the equivalent cylinder trength
of the concrete as it exists in the structure. If in-place tests
are u ed a valid relation hip between the re ults of in-place
tests and the compressive strength of cylinder is required.
I present, there are no standard practices for developing the
required relation hip.
1.4-Recommendations in other ACI documents
After the 1 995 ver ion of thi report wa published, other
I documents incorporated in-place test as alternative
procedures for estimating in-place strength. One of these docu­
ments is
I 30 I a specification for new concrete con truc­
tion. In the 20 1 6 ve1 ion of
I 30 I , ection 1 .6.4.2 on
in-place testing of hardened concrete include the following:
U e of the rebound hammer in accordance with
A TM C805/C805M or the pulse-velocity method
in accordance with ASTM CS97 may be speci­
fied by
rchitect!Engineer to evaluate unifor­
mity of in-place concrete or to elect areas to be
cored. These method shall not be used to evaluate
in-place strength.
Regarding the validity of in-place strength test , A I
30 1 -16 states in ection 1 .6.S.3(a):
Result of in-place strength tests will be evaluated
by Architect/Engineer 311d are valid only if tests are
condu ted u ing properly calibrated equipment in
a cordance with recognized standard procedure
and an acceptable correlation between te t re ult
and concrete compres ive trength i established
and submitted.
ection 1 .6.6.4 of
I 30 1 - 1 6, however, restricts the
u e of the e te ts in acceptance of concrete by stating
that ''In-place tests shall not be used as the sole basis for
accepting or rejecting concrete," but they may be used if
specified to "evaluate" concrete if standard-cured cyl inder
trength fail to meet the pecified strength criteria.
A I 30 1 - 1 6 al o mention in-place te t in ection 2.3.4
dealing with required strength for removal of formwork.
pecifically, it is stated that if specified in Contract Docu­
ments the following methods may be used to estimate
in-place trength:
73/ 873M (ca t-in-place cylinder )
(a)
TM
(b) A TM 803/C803M ( penetration resi tance)
(c)
TM 900 (pul lout)
TM C I 074 (maturity method)
(d)
The e same methods may be u ed if pecified, as alterna­
tive to testing field-cmed cylinder for e timating in-place
tr ngth for the purpo e of terminating curing procedure .
A J 562- 1 6, the repair code for exi ting concrete build­
ing , allowed the use of in-place test methods for assessment
of concrete trength. ection 6.4.3.2 state :
American Concrete Institute - Copyrighted © Material - www.concrete.org
4
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228. 1 R- 1 9)
ondestructive strength testing to evaluate in-place
strength of concrete shall be permitted if a alid
correlation is established with core sample compre i e trength te I re ult and nondestructive te t
re ult . Quantifications of concrete compre sive
strength by DT alone shall not be permitted as a
substitute for core amp l ing and te ting.
ACJ 308. 1 also mention in-pia e te t a acceptable
method for e timating in-place trength for the purpo e
of terminating curing procedures. Project specifications
can, therefore, reference standard specifications that allow
in-place testing as an alternative to testing field-cured cylin­
ders. A I 325. 1 1 R di cu ses the use of in-place te t for
e timating early-age concrete strength in fast-track concrete
paving projects.
In all cases, ufficient correlation data are requi red, and
permis ion may have to be granted before using in-place
test method . Thi report explains how the required correla­
tion data can be acquired and provide guidance on how to
implement an in-place te ting program.
CHAPTER 2-NOTATION AND DEFIN ITIONS
2.1 -Notation
a
b
c
d,_.
I
i
In
lnl
lnPO
K
k
M(t)
111
1 .5-Existing construction
Reliable estimates of the in-place concrete trength are
required for structural evaluation of exi ting structure ( C I
4 3 7 R and A I -62). H istorically, in-place strength has been
e timated by te ting core drilled from the tructure. In-place
tests can supplement coring and penni! more economical
c aluation of concrete in the structure. The critical step in
such applications is to establi h the relationship between
in-place te t re ult and concrete str ngth. More recently the
approach i to correlate re ult of in-place te t performed at
selected locations with the strength of corresponding cores.
Although in-place testing does not eliminate the need for
coring, it can reduce the total amount of coring needed to
evaluate a large volume of concrete.
ound sampling plan
is required to acquire the correlation data, and appropriate
stati tical methods should be u ed for reliable interpretation
o f te t re ult .
11;
n,
n,
PO
Q
=L(X-Xf
1 .6-Report objective
This report re iew
TM te t method for e timating
the in-place trength of concrete in new construction and
existing structures. The overall objective is to provide the
potential user with a guide to as ist in planning, conducting
and interpreting the result of in-place test .
hapter 2 inc ludes the notation and defi nition of terms
used in this report. hapter 3 discusses the underlying prin­
ciples and inherent limitation of in-place te ts. hapter
4 review the statistical characteri tic of in-place test .
hapter 5 outl ines procedures to develop the relation­
hip needed to e ti mate in-place compre i e trength.
Chapter 6 d i cu e factors to be con idered in planning the
in-place testing program. Chapter 7 pre ents statistical tech­
niques to interpret in-place test results. Chapter 8 discusses
in-place testing for acceptance of concrete. Appendix A
pro ides detail and an i llustrative example on the statistical
principles di cus ed.
(aCiJ
=
intercept of line
slope of line, u ed m development of trength
relationship
average of in-place compressive strengths
individual compres ive strength test results
de iation of each te t point from the best-fit l ine
in-place test result
indjvidual in-place trength te t re ults
average of natural logarithms of compre 1 e
trength
average of natural logarithms of in-place te t re ult
average of natural logarithms of pullout load
one-sided tolerance factor (Table 6.1 .2a)
b/'J..., where A. is obtained from the single-operator
variability during correlation te ting as given by
q. ( .2d).
temperature-time factor at age t. deg-days or
deg-hours
number of replicate in-place tests perfonned on an
element of a structure
number of te t point used to e tabli h the trength
relation hip
number of replicate in-place te ts
number of replicate in-place tests at each strength
level
number of replicate compressive strength tests at
each trength le el
average of natural logarithms of pullout loads
apparent activation energy di ided by the gas
con tant. K (kelvin)
estimated re idual tandard deviation
modified sum of the quares as given by Eq. (A.2i(a))
urn of quare of the deviation about X of the X
alue us:_d to develop the trength relationship, Sa
if
sx
tandard deviation of the logarithms of concrete
strength in the structure
tandard deviation of logarithm of compre si e
trength in the laboratory
error of fit of trength relation hip a given by Eq.
( .2j)
tandard deviation of logarithms of the in-place
re ult in the laboratory
tandard deviation of the logarithm of the in-place
te ts perfonned on the structure
standard deviation of the logarithms of the in-place
tests at strength levelj
standard deviation of estimated value of Y (average
concrete strength)
standard deviation of the logarithms of the compre ive strength test at strength levelj
average concrete temperature during time interval
r, oc (°F} or K
datum temperature, 0 (°F)
pecified curing temperature, K
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228. 1 R-19)
equivalent age at a specified temperature 'fs, days or
hours
t,.1.a
tudent's /-value for m- 1 degrees of freedom and
confidence le el a
t.v-2.o12 = tudent's /-value for -2 degrees of freedom and
ignificance level a
coefficient of variation of in-place test
Vi
V,
coefficient of variation of standard test
Vr
coefficient of variation (ex pre ed a a decimal)
W
width of the I 00( 1 - a) percent confidence interval
for the estimated mean value of Y for the value X
average of the logarithms of the in-place tests
X
perfonned on the structure
average of the logarithm of the in-place tests at
�
strength le elj
grand average of X values during coJTelation tests
X
q. (A.2g(a))
Y
logarithm of the e timated average in-place
compre i e trength
grand averag of Y value during correlation te t ,
Y
Eq. (A.2g(b))
lower tenth percentile of strength (I 0 percent
Yo.1o
defective)
average of the logarithms of the compress• e
>j
trength tests at strength level}
lower confidence limit of e timated average
Y10,.
in-place trength at confidence level a
time interval, day or hour
1
- ratio of variance of average compre sive strength
A.
te t re ults to ariance of average in-pia e test result
I�
2.2-Defi n itions
5
maturity inde x
an indicator of marurity, such as equi a­
lent age or temperature-time factor, that is calculated from
the temperature history of the cementitiou mixture by using
a maturity function.
residual tandard deviation-a measure of the scatter of
test data about a regres ion line· the tandard deviation of the
test value from the value c timated from the regre sion line.
single-operator v aria b il ity
the standard deviation
or coefficient of variation of repl icate test detenninations
obtained on identical test specimens by a single operator
using the arne apparatus.
strength rel a tionship--empirical equation obtained from
regression analy is of test data that relate the compressi e
strength of concrete to the re ult of an in-place test method.
temperature-time factor the maturity index computed
as the area between the concrete temperature and the datum
temperature from the plot of mea w-ed concrete temperature
crsu time, expre sed in unit of degree-days or degree-hours.
-
-
-
CHAPTER 3-REVIEW OF M ETHODS
3 . 1 -l ntroduction
ften, the objective of in-place testing is to estimate
the compre sive trength of concrete in the tructure. To
make a trength e timate, it i necessary to have a known
relation hip between the re ult of the in-place te t and the
concrete strength. For new con truction, thi relationship
i u ually e tabli hed empirically in the laboratory. For
exi ting con truction, the relationship is con entionally
e tablished by perf01ming in-place te t at selected loca­
tions in the structure and determining the strength of cores
drilled from adjacent location . Cl 2 1 4 4 R discu es the
number of core pecimens needed to obtain, at a given confi­
dence level, an e timate of average trength that is within a
certain percentage of the true average strength. Figure 3.1 is
a schematic of a strength relationship in which the cylinder
compressive trength is plotted a a function of an in-place
te t re uJt. Thj relation hip would be u ed to e timate the
tr ngth of concrete in a tructure ba ed on the value of the
in-place te t result obtained from te ting the structure. The
accuracy of the strength estimate depend on the degree of
correlation between the strength of concrete and the quan­
tity measured by the in-place test. The user of in-place tests
should understand what property i measured by the test and
how this property i related to the strength of concrete.
The purpo e ofthi chapter i to explain the underlying prin­
ciples of the widely used in-place te t methods, and to iden­
tify the factors other than concrete strength, that can influence
the te t re ult . Additional background information on the e
methods i available in the refi renee by Malhotra ( 1 976),
Bw1gey et al. (2006), and Malhotra and Carino (2004).
The fol lowing method are discus ed:
(a) Rebound number
(b) Penetration resistance
(c) Pullout test
(d) Pull-off test
(e) ltrasonic put e velocity
(t) Maturity
(g) a t-in-place cylinder
.
A I pro ide a comprehensive li t of definitions through
an online resource,
I oncrete Terminology. Definition
pro ided herein complement that ource.
apparent activation energy- an empirical factor that
indicate the temperature sensitivity of the rate of trength
development of a cementitious mixture after final setting has
occurred.
datum temperatur
the temperature value that is used
for calculating the temperature-time factor.
equivalent ag
the number of days or hours of curing
of a concrete mixture at a specified temperature required to
produce a maturity equal to the maturity achieved by a given
curing period at concrete temperatures different from the
specified temperature.
lower confidence limi t - the value above which the true
average, in-place strength i expected to occur with a speci­
fied probability or confidence level.
lower tenth percentile s t rength- the trength below
which I 0 percent of test results are expected to occur.
maturity-the extent of the development of a property of
a cementitious mixture.
maturity function a mathematical expression that u es
the measured temperature hi tory of a cementitious mixture
during the curing period to calculate an index that is indica­
tive of the maturity at the end of that period.
American Concrete Institute - Copyrighted © Material - www.concrete.org
6
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228. 1 R- 1 9)
(a)
Instrument ready
for test
Cyl i nder
Compressive
Strength
(b)
Body pushed
toward test
object
(c)
Hammer is
released
(d)
Hammer
rebounds
In-Place Test Value
Fig. 3.1
chemotic of relationship between cylinder
compressive strength and in-place test value.
3.2-Rebound number (ASTM C805/C805M)
Operation of the rebound hammer, also called the chmidt
Hammer, is illustrated in Fig. 3.2. The
Hammer or \ i
device consi ts of the following main component : I) outer
body; 2) metal plunger and guide rod· 3) hammer; and 4)
impact pring. The fol lowing equence de cribes how the tc t
i pcrfonned. From the initial locked condition, the plunger
i placed in contact with the concrete urface and the plunger
i extended from the body of the in trument by mean of a
compres ion pring. A latch lock the hammer in place and
the instrument i ready for testing ( Fig. 3 .2(a)). Then, the body
i pu hed toward the te t object, which extend the impact
spring connected between the hammer and body (Fig. 3.2(b)).
When the body i pu hed to its l imit of tra el, the hammer i
released and travels toward the test object due to the energy
tored in the tretched impact spring (Fig. 3.2(c)). Finally, the
hammer trike the plunger and rebounds a certain di tance
along the guide rod (Fig. 3 .2(d)). The rebound number can
be based on the rebound di lance expre ed as a percentage
of the tretch dj tance of the impact pring, or it can be ba ed
on the ratio of the hammer peed after impact to the hammer
peed before impact with the plunger.
The key to understanding the inherent limitations of thi
te t for estimating strength i recognizing the factors influ­
encing the rebound number. The impact loading and re ulting
wave propagation within the hammer-concrete sy tern re ult
in a compl icated dynamic sy tern that is difficult to model
from a fundamental point of view.
The rebound number depends on the kinetic energy in
the hammer before impact with the plunger and the amount
of that energy ab orbed during impact. Part of the energy
i ab orbed a mechanical friction in the in trument, and
part of the energy is ab orbed in the interaction of the
plunger with the concrete. The latter factor is what make
the rebound number an indicator of the concrete properties.
nergy ab orbed by the concrete depends on the tres - train
relationship of the concrete. Therefore, ab orbed energy i
related to the strength and stiffne of the concrete.
low­
strength, low- tiffnc s concrete will absorb more energy
than a high-strength, high-stiffnes concrete. Thu , the
low- trength concrete will likely re ult in a lower rebound
number. Because it i po ible for two concrete mixture to
have the arne trength but d ifferent st iffne e , there could
be d ifferent rebound number e en if the trengths are equal.
(aCiJ
Fig. 3.2- -S hematic illustrating operation ofrebound hammer.
Conversely, it i pos ible for two concrete with ditrerent
trength to have the arne rebound number if the tiffne
of the low- trength concrete i greater than the stiffne of
the high- trcngth concrete. Bccau c aggregate type and the
olume of coar e aggregate affect the stilfne of concrete, it
i nece sary to develop the strength relation hip on concrete
made with the arne aggregates and imilar proportions that
will be u ed for the concrete in the structure.
ln rebound-hammer te ting, the concrete near the point
where the plunger impact influence the rebound value.
Therefore, the test is ensitive to the urface condition , local
ariation in concrete con olidation, and relative tiffne
of the member at the location where the te t i performed.
If the plunger i located over a hard-aggregate particle an
unu ually high rebound number will re ult. By contra t, if
the plunger i located over a large air oid or a oft aggre­
gate particle, a lower rebound number will occur. Reinforcing
bars with hallow concrete co er could al o affect rebound
numbers if test are done directly o er the bars. To account
for these po sibilitie
TM C 051 05M require that 1 0
rebound numbers be taken for a test. I f a reading differs by
more than ix units from the average, that reading i di carded
and a new a erage i computed ba ed on the remaining read­
ing . I f more than two reading differ from the a erage by i
unit , the entire et of reading i di carded.
Becau e the rebound number i affected mainly by the
ncar- urfacc layer of concrete. the rebound number might
not repre ent the interior concrete. The pre ence of urface
carbonation can re ult in higher rebound numbers that
are not indicative of the interior concrete. imilarly, a dry
surface will re ult in higher rebound number than for the
moi t, interior concrete. Ab orptive oiled p lywood can
ab orb moi ture from the concrete and produce a harder
urface layer than concrete ca t again t teel form . imi­
larly, curing condition affect the trength and tiffne of
the near- urface concrete more than the interior concrete.
The urface tc turc can al o influence the rebound number.
When the te t i performed on rough concrete, local crushing
occur under the plunger and the indicated concrete trength
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228. 1 R-19)
will be lower than the true value. Rough surfaces should be
ground before te ting. I f the formed urfaces are smooth,
grinding is unneces ary.
hard, smooth surface, such as a
urface produced by trowel fini hing, can result in higher
rebound number . Finally the rebound di tance is affected
by the orientation of the instrument· the strength relationship
has to be developed for the arne instrument orientation as
will be u ed for in-place testing.
Ln wnmary, while the rebound number te t is imple to
perform, there are many factor other than concrete trength
that influence test re ults. As a result, estimated compre sive strengths are tmreliable unless a correlation is devel­
oped between rebound number and compressive strength for
a given concrete mixture or from core taken from a struc­
ture. Refer to hapters 5 and 6 for additional information on
developing the relationship and on using the relation hip to
e timate in-place strength.
3.3-Penetration resistance (ASTM C803/C803M)
In the penetration-re i lance technique, depth of pene­
tration of a rod (probe) or a pin forced into the hardened
concrete by a dri er unit i measured.
The probe-penetration technique involve the use of a
specially designed gun to drive a hardened steel probe into
the concrete. One well-known commercial test system is
the Wind or Probe. The penetration depth of the probe is an
indicator of concrete strength. This method is imilar to the
rebound number te t, except the probe impact the concrete
with much higher energy than the plunger of the rebound
hammer. The probe penetrates the concrete, compared with
the plunger of the rebound hammer which only produce a
minor urface indentation. A theoretical analysi of thi te t
i even more complicated than the rebound test, but again,
the e sence of the test invol es the initial kinetic energy
of the probe and energy absorption by the concrete. The
probe penetrates the concrete until it initial kinetic energy
i absorbed. The initial kinetic energy i governed by the
charge of smokeles po' der used to propel the probe the
location of the probe in the gun barrel before firing, and
frictional lo ses as the probe travels through the barrel. An
e sential requirement of thi test is that the probe have a
con i tent alue of initial kinetic energy. A TM C 03/
03M require that the probe exit-velocitie do not ha e
a coefficient of variation greater than 3 percent based on I 0
tests by appro ed balli tic methods.
As the probe penetrates the concrete, some energy is
absorbed by friction between the probe and the concrete, and
some is ab orbed by cru bing and fracturing the concrete.
There are no rigorous tudies of factor affecting the geom­
etry of the fracture zone, but it general shape i mo t likely
similar to Fig. 3.3a. There i usually a cone-shaped region
in which the concrete i heavily fractured and a ignificant
pot1ion of the probe energy i absorbed in thi zone.
The probe tip can travel through mot1ar and aggregate;
in general, cracks in the fracture zone will be through the
mortar matrix and the coarse-aggregate particles. l -Ienee,
the strength properties of both the mortar and coarse aggre­
gate influence penetration distance. This contra t with the
7
Exposed
length
f'S·.::•.··,\iiii�� Coarse
aggregate
Fig. 3.3a-Approximate shape offailure zone in concrete
during probe penetration test.
Penetration Depth, in.
04
l1l
0...
�
.s::
0,
c::
Q)
....
U5
Q)
>
"iii
(/)
Q)
....
a.
E
0
u
08
1 2
1 6
2
5800
40
"iii
a.
30
4350
2900
20
. . . . . . . . . -� . . . . . . . . . . . . . i . . . . . . . . . .
10
0
10
..
..
..
..
.
..
..
..
..
.
20
30
1 450
.s::
0,
c::
Q)
....
....
(f)
Q)
>
"iii
(/)
�
a.
E
0
(.)
0
40
50
Penetration Depth, mm
Fig. 3. 3b-Effect of aggregate type on relation hip between
concrete strength and depth ofprobe penetration.
behavior of normal- trength concrete in a compre sion te t,
where mortar trength has the predominant influence on
mea ured compre ive trength. Thu , an important charac­
teri tic of the probe penetration te t i that the type of coar e
aggregate affect greatly the relation hip between concrete
trength and depth of probe pen tration. For example Fig.
3.3b compare empirical relationship between compre ive trengtb and prob penetration for con rete made with
a oft aggregate uch as Lime tone, and concrete made
with a hard aggregate uch a chert. For equal cornpre si e
trength , the concrete with the oft aggregate allow greater
probe penetration than concrete with hard aggregate. More
detailed information on the influence of aggregate type on
trength relation hip i found in Malhotr. ( 1 976), Bungey
et al. (2006), and Malhotra and arino (2004).
Penetration value may be affected by urface v ith coar e
texture . ASTM C803/C 03M require grinding of urfa e
that are coar er than burlap-dragged fini be before te ting.
hard urface layer, a would o cur with trowel finishing,
can re ult i n low penetration values and exces ive scatter of
data. The orientation in which the te t i performed i not
critical provided the probe is driven perpendicular to the
American Concrete lnstitu1e - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228. 1 R- 1 9)
8
surface. The penetration, however will be affected by the
pre ence of reinforcing bars within the zone of influence of
the penetrating probe. Thus, the location of the reinforcing
tee) should be determined before electing test ites. over
meter can be used for thi purpo e (
I 228.2 R).
In practice, it is cu tomary to measure the exposed length of
the probe . The fundamental relationshjp however, i between
concrete trength and resistance to penetration. Therefore
when a se sing the ariabi)jty of te t re ult (Chapter 4), it i
preferable to expre the coefficient of variation in tenn of
penetration depth rather than expo ed length.
Before 1 999, the hardened steel probes were limited to
� e in concrete with compre sive strength less than approx­
Imately 40 M Pa ( 6000 p i ) . There wa a tendency for the
probe to fracture > ithin the threaded region when testing
stronger concrete. Al-Manaseer and qui no ( 1 999) reported
that a newer probe made with tre s-relieved alloy tee)
was successfully used to test concrete with a compres ive
strength of 1 1 7 M Pa ( 1 7,000 p i).
A pin penetration te t device, which u e le
energy
than the Wind or Probe y tem, wa developed by a ser
( asser and Al-Manaseer 1 987a,b), and the procedure for
its use was sub equently incorporated into A TM C 031
C803 M .
spring-loaded device i u ed to dri e a pointed
3.56 mm (0. 1 40 in.) diameter hardened tee! pin into the
concrete. Penetration by the pin creates a mall indentation
(or hole) in the concrete urface. The pin i removed from
the hole, the hole cleaned with an air jet, and the hole depth
mea ured with a uitable depth gauge. Penetration depth is
u ed o e timate compre si e strength from a pre iously
e tabh bed trength relation hip.
The kinetic energy deli ered by the pin penetration device
i � t i mated to be approxjmately 1 .3 percent of the energy
dell ered by the Windsor Probe system (Carino and Tank
1 989). Becau e of low energy le el, penetration of the pin
_ reduced greatly if the pin encounters a coarse-aggregate
IS
particle. Thu , the test is intended a a penetration test of
the mortar fraction of the concrete. Results of tests that
penetrate coar e-aggregate particles are not con idered in
detennining the average pin-penetration resistance (ASTM
C803/C803 M). A pin can b come blunted during penen·a­
tion. Becau e the degree of blunting affe t penetration
depth,
TM C803/C803M require that a new pin be u ed
for each penetration test.
ensiti ity of the pin penetration to changes in compressive
strength decreases for concrete strength above 2 M Pa (4000
psi) (Carino and Tank 1 9 9). Therefore the pin penetration
te t ystem i not recommended for testing concrete having
an expected compre ive n·ength above 2 MPa (4000 p i ).
In summary concrete strength can be estimated by
mea uring the penetration depth of a probe or pin driven
into the concrete at con tant energy. Penetration te t are
le affected by urface condition than the rebound number
method. The coar e aggregate, however ha a sigruficant
effect on the re ulting penetration. For the gun-driven probe
system, the type of coar e aggregate affects the strength rela­
tionship; for the spring-driven pin system tests that impact
coar e aggregate particle are di regarded.
�
(aCiJ
d, <: 1 .25 d,
20
d, < d, < 2 4 d,
,.
'
Fig. 3. 4a-Dimensiona/ requirements of pullout test in
accordance with A TM 900.
3.4-Pullout test (ASTM C900)
The pullout test measures the maximum force required to
pull an embedded metal insert with an enlarged head from
a concrete specimen or structure. A po itivc feature of the
pullout te t i that it produce a well-defined failure urface
in the concrete and mea ure a tatic, rather than dynamic,
strength property of the concrete.
The test requires a metal insert that is either cast into fresh
concrete or installed into hardened concrete. The pullout
force i s applied by a loading system that react against the
concrete surface through a reaction ring concentric with
the in ert ( Fig. 3 .4a). A the insert is pulled out, a roughly
c � ne-shaped fragment of the concrete is extracted. The large
d1amcter of the conic fragment, d2, i detennined by the
inner diameter of the reaction ring, and the mall diameter
d1 i dctennined by the insert-head diameter. Requirement
for the te ting configuration are given in
TM 900. The
embedment depth and head diameter have to be qual, but
there i no requirement on the magnitude of the e dimen­
sion . The inner diameter of the reaction ring can be between
2.0 and 2.4 times the insert-head diameter. This means that
the apex angle ("2a") of the conic frustum defined by the
in ert-head diameter and the inside diameter of the reac­
tion ring can vary between 54 and 70 degree . The arne test
geometry ha to be used for de eloping the trength relation­
ship and for the in-place testing.
Unlike the rebound hammer and probe-penetration te t , the
pullout te t ubject the concrete to a tatic loading that lend
it elf to tre analy is. The finite-element method ha been
u ed to calculate the stre e induced in the concrete before
cracking ( tone and arino 1 984) and where the concrete has
cracked ( ttosen 1 9 I · Ferretti 2004). In the e analyses the
concrete was as umed to be a homogeneous solid and the influ­
ence of discrete coarse-aggregate particles was not modeled.
There is agreement in cited literature that the test subject the
concrete to a nonuniform, three-dimensional state of stres .
Figure 3.4b show the approximate directions (trajectorie )
of the principal tre e acting in radial plane (tho e pas ing
through the center of the in ert) before cracking for apex
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228. 1 R-19)
9
Tensile stress trajectory
Compressive stress trajectory
Fig. 3. 4c-Circumferential cracks predicted by nonlinearfrac­
ture mechanics analysis ofpullout test by Hellier et a!. {1987).
--
- - - - ·
Tensile stress trajectory
Compressive stress trajectory
Fig. 3. 4b-Prin ipal stre s trajectorie before cracking for
pullout test in a homogeneous material and measured frac­
ture swfaces in physical tests (Stone and Carino 1 984).
angles of 54 and 70 degrees. Becau e of symmetry, only one­
half of the pecimen i shown. The e trajectories would be
expected to change after cracking de elop . Before cracking,
there are ten iJe stre e that are approximately perpendicular
to the eventual failure surface mea ured by Stone and Carino
( 1 984). Compressive tre es are directed from the insert head
toward the ring. The principal stresses are nonuniform and are
greatest near the top edge of the in ert head.
A serie of analytical and experimental studies some of
which are critically reviewed by Yener and hen ( 1 984
has been carried out to detennine the fai lure mechanism of
the pullout test. While the conclu io11s have been different, it
i generally agreed that circumferential cracking (producing
the failure cone) begin in the highly tres ed region next to
the in ert head at a pullout load that i a fraction of the ulti­
mate value. With increasing load, the circumferential cracking
propagates from the in ert head toward the reaction ring. There
,
is no agreement on the nature of the :final failure mechani m
governing the magnitude of the ultimate pullout load.
Ottosen ( 1 9 1 ) concluded that fai lure i due to "crushing"
of concrete in a narrow band between the insert head and
the reaction ring. Thu , the pullout load is related directly
to the compre ive trength of concrete. In another analyt­
ical tudy Ycner ( 1 994) concluded that failure occurred by
outward cmshing of concrete around the perimeter of the
failure cone near the reaction ring. Using linear-elastic frac­
ture mechanic and a two-dimensional mode� Ballarini et al.
( 1 9 6) concluded that ultimate pullout load is governed by
the fracture toughness of the matrix. In an experimental study,
tone and Carino ( 1 983) concluded that before ultimate load,
circumferential cracking extends from the insert head to the
reaction ring, and that additional load is re isted by aggregate
interlock aero the circumferential crack. In thi ca e, failure
occurs when ufficient aggregate particle have been pulled
out of the mortar matrix. According to the aggregate inter­
lock the01y, maximum pullout force i not directly related to
the compre i e trength. There i good correlation, however,
between ultimate pul lout load and compre i e trength of
concrete becau e both value are influenced by the mortar
trength (Stone and Carino 1 984). In other studies, using
nonlinear fracture mechanics and discrete cracking models,
Hellier at al. ( 1 987) and Ferretti (2004) showed excellent
agreement between the predicted and ob erved internal
cracking in the pullout te t. Figure 3.4c shows the displaced
hape of the finite-element model used by Hellier et al.
( 1 987). The analy e by Hellier et al. and Ferretti indicate that
a plimaty circumferential crack de eloped at the corner of
the in ert head and propagated outward at a shallow angle.
Thi crack cea ed to grow when it penetrated a ten i le-free
region. A secondary crack developed ub equently and prop­
agated as shown in the figure. The econdary crack appears
American Concrete Institute - Copyrighted © Material - www.concrete.org
10
REPORT ON METHODS F O R ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228. 1 R-19)
Flat Surface
h
=
d2
Core
hole
"'
d,
Expandable
insert
Expansion
tool
(a) Drill hole, plane surface, and undercut
(b) Insert expansion tool and expandable
slot
msert
(c) Expand tnsert
(d) Install bearing ring and loading system,
pullout insert assembly
Fig. 3.4d- Techniquefor post-in tailed pul/out te I (adaptedfrom ASTM C900).
to coincide \ ith the final fra ture urface ob er ed when the
conical fragment i extracted from the concrete ma during
pul lout te ting. The e tudie al o concluded that the ulti­
mate pul lout load is not governed by unia ial compre i e
fai l ure in the concrete.
Pul lout strength i primarily governed by the concrete
located next to the conic fru tum defined by the in crt head
and reaction ring.
ommercial in crt ha e embedment
depth of appro imately 25 to 0 mm ( 1 to 1 .2 i n . ). Thu ,
only a mal l olumc of concrete i tc ted and, becau c of
the inherent heterogeneity of concrete, the average ingle­
opcrator coeffi ient of variation of the e pul lout te t h
been found to be between approximately 4 and 1 5 percent.
In ne> con !ruction, the mo t de irable approa h for
pullout testing is to attach the in crt to formwork before
concrete placement. It i a! o po ible, howe er, to place
in crt into unformed urface , uch a top of slab , by
placing the in crt into fresh concrete that i ufficiently work­
able. The hardware incl ude a metal plate attached to the in crt
to pro ide a mooth bearing urface, and a pia tic cup to al lo�
embedment of the plate lightly bclo> the urfacc. The pia tic
cup al o en urc that the in crt ' i l l float in the fre h concrete
and not ettle before the concrete set . I f insett are placed
manually, care i required to maintain repre entative con rete
propertie at placement location and to reduce the amount of
air that i entrapped on the under ide of the plate . In an early
study, Vogt et at. ( 1 9 4) reported higher-than-exp cted ingle­
operator variabi lity \ hen using manually placed in crt .
Sub equent work by Dil ly and Vogt ( 1 98 ) re ulted in an­
abi l ity imi lar to that expected with insert fastened to form­
work. The recommended approach is to pu h the insert into
fre h concrete and then float it horizontally o er a distance of
50 to I 00 mm (2 to 4 in.) to al low aggregate to Aow into the
pullout fai lure zone. fter in ertion, the in crt hould be tilted
about 20 to 0 degree from the vertical to allow entrapped air
to c cape from beneath the tcel plate. Care hould be taken to
en urc that the plate i completely belo the con rete urface.
To prevent mo ement of the in crt before the on rete et ,
fre h concrete can be placed in the cup.
In e i ling con truction, it i po ible to perform pul lout
te t
u ing post-in tailed insert . The procedure for
performing po t-in tai led pul lout te ts, which \ a added to
A TM C900 in 1 999, i summarized in Fig. 3 .4d. The proce­
dure involve five ba ic tep :
( I ) Dri l l ing a hole perpendicular to the urface of the
concrete u ing a core dril l
( 2 ) Grinding the t e t area t o make i t flat
( 3 ) Undercutting a lot to engage an e pandablc in crt
(4) Expanding an in crt into the mil led lot
( 5 ) Pull ing the in crt out of the on crete
The te t geometry i the ame as for the ca t-in-place
in crt. In a commercial te t sy tern known a the APO
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228 . 1 R - 1 9)
(Cut and Pull Out) test the insert is a coiled, split ring that
i expanded with specially designed hardware. The CAPO
system perfonns similarly to the cast-in-place system of
the arne geometry ( Petersen 1 9 4, 1 997). are i required
during preparation to ensure that the hole i drilled perpen­
djcular to the test surface. The urface has to be flat o that
the bearing ring of the loading sy tem is supp01ted unjfonnly
when the in ert i extracted. onunifonn bearing of the
reaction ring can re ult in an incomplete circle for the top
urface of the extracted frustum. l f thi occur , the te t re ult
mu t be rejected (A TM C900). Water u ed for drilling and
undercutting should be removed from the hole a soon as the
undercutting is completed, and the hole should be protected
from ingres of water until the te t is completed. This is to
prevent penetration of water into the fracture zone, which
might affect the measured pullout load.
Other types of pullout te t configmation are a ailable
for existing construction ( Mai lhot ct al. 1 979· habowski
and Bryden-Smith 1 980; Dom n and Castro 1 9 7). The e
typically in olve drilling a hole and in erting an expanding
anchorage device that will engage in the concrete and cau e
fracture in the concrete when the device is extracted. The e
methods, however do not have the same fai lure mechanisms
as the standard pullout test. These techniques have not been
tandardized as A TM te ts methods; howe er, the internal
fracture test by habow ki and Bryden- mith ( 1 980) has
been incorporated into a British standard ( B 1 8 1 -Part 207).
In ummary. the pullout test can be u ed to estimate the
strength of concrete by measuring the force required to
extract an insert embedded in fresh concrete or installed
in hardened concrete. The test re ult in a complex, tbree­
dimen ional state of stres in the concrete. Although there i
no consensu on the exact failure mechanism of the pullout
test there is a strong relationship between the compressi e
strength and pullout . trength of concrete.
3.5-Pul l-off test (ASTM C1 583/C1 583M)
The pull-off test as de cribed in A TM 1 5 3/C I 83 M,
i u ed to detennine the near- urface ten ile trength of the
substrate as an indicator of the adequacy of surface prepa­
ration before application of a repair or o erlay material;
the bond trength of a repair or an overlay material to the
ub trate; and the tensil strength of a repair or overlay
material, or an adhesi e used in repairs, after the material
has been applied to a surface (Long and Murray 1 984). The
pull-off method is illustrated in Fig. 3.5. In the A TM test
procedure, the surface to be tested is cored to a depth of at
lea t I 0 mm (0.5 in.), as shown in Fig. 3 .5(a). If testing the
bond trength of an overlay or a concrete repair the core
depth is extended to at lea t I 0 mm (0.5 in.) below the
concrete overlay inte1face, as hown in Fig. 3 .5(b).
metal
dj c i then bonded to the te t urface u ing a uitable adheive. A temporary packing material can be placed in the kerf
to prevent the adhe ive from fil ling in the annular cut. The
ruse is then pulled from the stuface at a con tant rate of 35
± 1 5 kP
(5 :r: 2 psils) using a device that ensure the force
i parallel to and coincident with the axis of the core. The
system to provide reaction to the applied load can be a ring
11
Tensile load1ng
deYIOe
Swivel JOtnt
Partial depth core at
least 10 mm (0.5 1n )
Adhesrve
into c:oocrete
Base concrete
(a) Test of base concrete
P rtllll depth core
through overlay or
repalf matenal and
least 10 mm (0 5 '" )
into base concrete
Base concrete
(b) Test of material applied to base concrete
(i) Faiture In
base concrete
(iQ Bond failure at
011ertay/ccncrete
lntelfaal
(IV) Disc bond
rllure
(c) Potential failure locations
Fig. 3. 5-Schematic of pull-off te t and potential failure
location (adaptedfrom ASTM 1 583/C / 583M).
or tripod. ourard et al. (20 1 4) invc tigated the influence of
pull-off force mi alignment and concluded that a 4-degrec
mi alignment could redu e th mea ured pull-off trength by
appro imately 1 5 p rcent.
An alternative procedure (not tandardized a of 20 1 6) i
to first prepare a flat surface u ing a planing tool that \ orks
without water, remove the du t, bond the disc, and then cut
American Concrete Institute - Copyrighted © Material - www.concrete.org
12
REPORT ON METHODS F O R ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228 . 1 R - 1 9)
the partial-depth core. This avoids waiting for the surface to
dry after cutting the core and before bonding the disc. This
procedure, however, require a core barrel that is properly
ized for the di c being used o that the partial-depth core is
centered with the disc.
The failure location in the pull-off test will be in the mate­
rial with the lowe t strength. As hown in Fig. 3.5(c), the
fai lure location can be at the concrete urface (dis bond
fai lure), at the interface between materials (bond fai lure), in
the overlay material (overlay fai lure), or in the ba e mate­
rial ( ubstrate failure). Because the failure location carmot
be predicted before the test is complete, both the fai lure load
and location are recorded after completion of the test. The
a erage strength in the pull-off te t is based on test re ults
with the same failure location.
The pull-off test was initially developed at Queen's
Uni ersity Belfa t in the 1 970s for strength testing of
in-place concrete. Initial research resulted in the develop­
ment of L I M PET te t equjpment ( Bungey and Madandoust
1 992 ), although everal type of commercial equipment are
a ailable. Early research by Bungey and Madandoust ( 1 992)
examined the effect of the disc material disc size, effect
of coring depth, reaction system, and loading rate. These
factors were examined in a series of experimental studies
and finite-element simulation . Thi research howed that
te t re ult are ensitive to different disc material (steel or
aluminum), di c diameter and thickne s, and core depth.
Experimental result showed that pull-off strength \ ere
not influenced by whether the loading system used a ring or
tripod arrangement to provide reaction to the applied load.
The re ult indicated that a 50 mm (2 in.) d iameter di c with
a di c thickne of30 mm ( 1 .2 in.) will produce a reasonably
uniform concrete stress distribution regardle of the ela tic
properties of the concrete or disc material.
TM 1 583/
1 5 3M requ i res a steel disc diameter of 50 mm (2 in.) with
a thickne s of at least 25 mm ( I in.).
The pull-otr te t has been u ed widely in evaluating the
bond strength between material in repair applications { I RI
2 1 0.3 R). In this application, the test can be used to a es the
surface preparation quality a well as bond trength. between
a repair material and ub trate. The I Rl reference provides a
de cription of the variou failure mechani m that can occur
when used in thi application. The pull-off test wa u ed
extensively to asses the effect of different concrete removal
or surface preparation methods on the bond of repair mate­
rials to the concrete substrate { I I indo 1 990). The research
by H indo identified that "bru ising " or micro-fracturing, of
the sub trate surface layer during concrete removal has a
significant effect on the bond strength of repair material ,
with lower bond strength observed with more aggre i e
removal methods. The more aggre sive remo al methods
re ult in damage to the concrete ub trate, often re ulting in
bond fai lure immediately below the interface between the
repair material and ubstrate.
Results from three different pull-off test device were
compared by
ay hurd and McDonald ( 1 999). Two
commercially available pull-off devices and a modified
device were used in the research. The modified device was
a commercially available instrwnent designed to asses the
pull-out strength of adhesive anchors and had been modi­
fied for pull-off testing. Pull-off tests were performed on a
concrete pecimen that included a imulated concrete repair.
The re earch program included an examination of the effects
of coring depth, with linear elastic, finite-element analy es
performed to examine the stre s concentration that occur
with different coring depth . Test results from the two
comm rcially available pull-off te t device were found to
be i m ilar, with lower pull-off trength obtained from the
modified device. The overall re ults of the pull-off tests had
high variability, with a preferential fai lure location observed
in the concrete substrate, rather than in the repair material
or at the bond line between materials. The result obtained
in this study were used for the preci ion tatement in
TM
1 583/ 1 583M (4.2 .4).
In summary the pull-off test can be u ed to a es the
bond between new materials and a concrete substrate, and
for qual ity of urface preparation before overlay placement.
Ba ed on the nature of the te t, both the failure location and
fai lure load are u d to asse s the test resul t . Te t re ult can
be affected significantly by surface preparation methods and
test sp cimen geometry.
3.6-Uitrasonic pulse velocity (ASTM C597)
The ultra onic pulse velocity te t, as prescribed in
TM
597, determine the propagation velocity of a pulse of
compres ional {P-wave) energy through a concrete member
(Jonc 1 949; Lc lie and Chcc man 1 949). The operational
principle of modern te ting equipment is il lustrated in Fig.
3.6a. A pul er ends a hort-duration high-voltage ignal to
a tran d ucer, causing the tran ducer to ibrate at it re onant
frequency and generate a stres wave pul e in the concrete
(refer to
I 228.2 R for a description of stress waves in
ol ids). At the start of the electrical pulse, an electronic t i mer
i witched on. The transducer vibrations are tran ferred to
the concrete through a viscou coupling fluid. The pul e
travels through the member and is detected by a receiving
tran ducer coupled to the opposite concrete surface.
hen
the pulse is received, the electronic timer is turned off and
the elap d travel time is dj played. The direct path length
between the tran ducer i divided by the travel tim to
obtain the pul e velocity through the concrete.
From the principles of elastic wave propagation, the
P-wave elocity is proportional to the square root of the
elastic modulu (A I 228.2R). Because the elastic modulus
and strength of a given concrete increase with maturity, it
follows that pulse veloci ty can provide a means of e timating
trength of concrete, even though there i no direct phy ical relation hip between these two properties.
concrete
mature , however, the elastic modulu and compre si e
trength increa e at different rate . t early maturities, the
ela tic rnodulu increa e at a higher rate than trength, and
at later maturities, the elastic modulu increa es at a lower
rate. A a result, over a wide range of maturity, the relationhip between compre sive trength and pul e velocity is
highly nonlinear. Figure 3 . 6b shows a typical relationship
between compressive trength and pulse velocity. ote that
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228. 1 R-19)
· - - - - - - - - - - - - - - - - - - - - - - - - - - - - -I
1
I
I
I
6560
8200
Pulse Velocity, ft/s
9840
1 1480
13
13120
� t-����������������&00
P ulser
ca
30
a..
�
25
0>
c
20
5
�
4000
..,..,
I!?
a.
E
0
u
£
Cl
c
3000 �
1/)
-�
'iii
a.
15
2000
10
1/)
Cll
>
·�
I!?
a.
E
0
1000 u
5
25
3
35
Pulse Velocity, km/s
4
Fig. 3. 6b-Schematic of typical relationship between pul e
velocity and compre sive trength ofa given concrete mixture.
Fig. 3. 6a- chematic of apparatu. to mea ure ultrasonic
pulse l'elocity.
this is only an il lustrative example and the actual relationhip depend on the pecific concrete mixture. At early
maturities, a gi en increase in compre sive strength is
accompanied by a relatively large increase in put e velocity,
while at later maturitie , the velocity increase i maller for
the arne trength increa e. For example, a trength increa e
from 3 to MPa (400 to 1200 p i roughly) can be associ­
ated with a velocity increa e from approximately 2400 to
3040 tnls (7900 to I 0,000 ftls roughly). However, a trength
increase from 25 to 30 MPa (3600 to 4400 psi roughly) could
correspond to a velocity increa e of only 3800 to 3920 m/s
( 1 2,500 to 1 2,900 ftls roughly). Thu , the sensitivity of the
put e elocity as an indicator of change in concrete trength
decrease with increasing maturity and strength.
Factors other than concrete trength can affect put e
elocity, and changes in put e velocity due to the e factor
could over hadow changes due to trength ( turmp et al.
1 984). For example, although the pulse velocity depends
strongly on the type and amount of aggregate in the concrete,
the strength of normal-strength concrete ( les than approxi­
mately 40 MPa [6000 psi ] ) is less sensiti e to the e factors.
As the volumetric aggregate content of concrete increa es
the pulse velocity may increase or decrease, depending on
the aggregate type, but the compre ive strength may not be
affected appreciably ( Boga et al. 2013; Jones 1 962).
Another factor i moi ture content. A the moi ture content
of concrete increa e from the air-dry to saturated condition,
put e elocity could incr a up to 5 percent (Bungey et al.
2006). I fthe effects of moisture are not considered, erroneous
conclu ions can be drawn about in-place strength, especially
in mature concrete. The curing proce s also affects the rela­
tionship between pulse velocity and strength, especially
when accelerated methods are used (Teodom 1986 ).
The amount and orientation of steel reinforcement will
at o influence pulse velocities. Because the pulse velocity
through steel is approximately 40 percent greater than
through oncrete, th pulse velocity through a heavily reinforced concrete member could be greater than through one
with little reinforcement. Thi i e pccially trouble orne
when reinforcing bar are oriented parallel to the pul e­
propagation direction. The pul e may be refracted into the
bar and transmitted to the recei er at the pul e elocity in
teel. The resulting apparent elocity through the member
will be greater than the actual clocity through the con rete.
Failure to account for the pre en c and orientation of rein­
forcement may lead to incorre t conclu ion about concrete
trength. Although con·ection fa tor have been propo ed,
uch a tho e di cu ed in Malhotra ( 1976) and Bungey et al.
( 2006), their accura y ha not been establi hed conclu ively.
The mea ured pulse velocity can al o be affected by the
pre ence of crack or oids along the propagation path from
!Tan miller to recei er. The put e could be diffracted around
the di continuitie , thereby increa ing the travel path and
ITa el time. Without additional knowledge about the interi r
condition of the concrete member, the apparent dccrca c
in put e velocity could be incorrectly interpreted
a low
ompre i e strength.
In thi te t method, all the con rete between the tran mit­
ling and receiving transducer affect the tra el time. Te t
re ult are, therefore, relatively in ensitive to the normal
heterogeneity of concrete. on equently, the te t method
has been found to ha e a low ingle-operator coefficient o f
ariation. Thi doe not mean, however, that trength e ti­
mate are neces arily highly rel iable.
TI1e u e of the velocity of other ultra onic wa e mode ( for
example, urfacc wave [Gudra and tawi ki 20 0; Gallo and
PopO\.ic 2005] ) and other wave chara tcri tic ( for c ample,
P-wa e wa e energy attenuation [damping] [Teodoru 1988;
I mail et al. 1996· Te famariam et al. 2006]) to e timate in-pia e
concrete trength have at o been reported. The application of
the e methods to concrete tmcture ha not, however, been
evaluated e tensively; furthermore, standard testing proce­
dure for these method have not yet been de eloped.
Tn summary, pulse velocity can be u ed to e timate trength
in new and exi ting con !ruction, pro ided a trength-rela-
American Concrete Institute - Copyrighted © Material - www.concrete.org
((]'CiJ
14
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R - 1 9)
tionship for the concrete mixture has been developed. or a
given concrete, a change in P-wave velocity is fundamen­
tally related to a change in ela tic modulus. Because elastic
modulus and trength are not linearly related pulse velocity
i inherently a less-sensitive indicator of concrete strength as
strength increases. The amount and type of aggregate has a
strong influence on the pul e elocity versus strength relationhip, and the in-place pulse velocity is affected by moisture
content and the pre ence of t el reinforcement. Refer to
I
228.2R for additional di cu sion of the pulse velocity method.
3.7-Maturity method (ASTM C1 074)
Freshly placed concrete gains strength because of the
exothennic chemical reactions between the water and
cementitious material in the mixture. Provided sufficient
moisture is present, the reaction rate are influenced by the
concrete temperature; an increa e in temperature causes an
increa e in the reaction rates. The extent of the reactions and,
therefore, strength at any age depend on the thermal hi tory
of the concrete.
The maturity method (
TM
I 074) i a technique to
estimate in-place strength by considering the relationship of
temperature and time on strength development. The thermal
history of the concrete and a maturity function are used to
calculate a maturity index that quantifies the combined effects
of time and temperature. The strength of a specific concrete
mixture i expressed a a function of its maturity index by
means of a strength-maturity relationship. If samples of the
same concrete are subjected to different temperature condi­
tions, the tr ngth-maturity relation hip for that concrete and
the temperature hi tories of the ample can be used to e ti­
mate their trengths.
The maturity function is a mathematical expression that
converts the temperature history of the concrete to a matu­
rity index. everal such functions have been proposed and
are reviewed in Malhotra ( 1 97 1 ), R I L M
omm1s 1on
42-CE ( 1 98 1 ) and Carino (2004). As explained by Carino
(2004) the maturity function i related to the rate of strength
development after fu1al setting has occurred. Therefore, a
key feature of a maturity function is the expression used to
represent th influence of temperature on the initial rate of
trength development. Two expres ion are commonly u ed.
I n one approach, it is as umed that the initial rate of strength
development is a linear function of temperature, which leads
to the imple maturity function, commonly known as the
urse- au I function
ur e 1 949; au I 1 95 1 ) shown in Fig.
3.7a. In this ca e the maturity index at any age is the area
between a datum temperature T0 and the temperature curve
of the concrete. The tenn temperature-time factor i u ed for
this area and is calculated as follows
M(t)
=
L.( Tu - To)!J.t
(3.7a)
where M(t) is temperature-time factor at age t, deg-days or
deg-homs; t::..t i a time interval, days or hours; To is average
concrete temperature during time interval tJ.t, 0 (°F); and T0
is datum temperature, °C (°F).
Temperature
T
To
Time t
Fig. 3. 7a-Maturity function based on assumption that the
initial rate ofstrength gain varies linearly with temperature;
shaded area is the temperature-timefactor (Eq. (3. 7a)).
Traditionally, the datum temperature u ed in Eq. (3.7a)
has been taken a the temperature below which tTength gain
cea es which ha been a umed to be approximately -1 0°C
( I 4°F). Tt ha been ugge ted, however, that a ingle value for
the datum temperature i not the mo t accurate approach and
that for better a curacy, the datum temperature hould be eval­
uated for the pecific cementitiou material and admixture
in the concrete mixtw-e (Carino 1 984). ASTM C l 074 recom­
mends a datum temperature of 0° (32°F) for concrete made
with A TM Type I ement when the concrete temperature i
e peeled to be between 0 and 40°C (32 and 1 04°F). A TM
I 074 provides a procedure to determine experimentally the
datum temperature for other type of cementitious material
and for different range of curing temperature.
Tn the econd approach it i a umed that the initial rate
of trength gain aries exponentially with concrete tempera­
ture in accordance � ith the AJThcniu function ( Frcic !eben­
Han en and Pedersen 1 977 ). Thi second maturity function
i u ed to compute an equivalent age of the concrete at a
specified curing temperature a follows
(3.7b)
where 1, i equivalent age at a pecified temperature T,., day
or hours; Q i apparent acti ation energy di ided by the ga
con taot, K (kelvin)· To i average ab olute temperature of
concrete during time interval t, K· T; i specified curing
temperature, K; and t is time inter al, days or hours.
In Eq. (3.7b), the exponential function can be considered
as an age conversion factor that convert a time interval t at
the actual concrete temperature to an equivalent interval (in
tenn of trength gain) at a pecified curing temperature. I n
orth America, the peci:fied curing temperature i typically
taken ro be 23°C (296 K), whcrca in Europe, 20°C (293 K)
i typically u ed. To calculat the equivalent age of a concrete
mixture, the alue of a parameter (known a the apparent
acti ation energy) i required; thi depends primarily on the
types of cementitiou material and, to a le er extent on the
water-cementitiou materials ratio (w/cm) ( arino and Tank
1 992). The Q-value in Eq. (3.7b) is the activation energy
divided by the ga constant (8.3 1 joules/(mole· K)). A TM
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
1 074 recommends a Q- alue o f 5000 K for concrete made
with ASTM Type I cement without any admixtures or addi­
tions and provides procedures for determining the Q-value
for other cementitious y terns. A Q-value of 5000 K corre­
sponds to an apparent activation energy of about 42 kJ/mol.
Figure 3 .7b show how the age conversion factor varies with
concrete temperature for different Q-values and a specified
curing temperature of23°C (73°F). As the Q-value increa e ,
the relation hip betw en age conver ion factor and tempera­
ture become more nonlinear. Researcher have obtained
apparent activation values for various cementitious systems
(Carino 2004· Brooks et al. 2007 ). Kjellsen and Detwi ler
( 1 993 ) suggested a modified version of the equivalent age
function to account for the observation that the apparent
activation energy decrease a hydration progres es, but
indicated that their model required further validation.
To use the maturity method requires establ ishing the
strength-maturity relation hip in the laboratmy for the
concrete that will be used in the tructure. A de cribed in
A TM C I 074, this i accomplished by preparing concrete
specimens (cylinders, beam , or cube ) to develop the
trength gain relationship in the laboratory, placing tempera­
ture sensors in two specimens, and measuring the strength
(compressive or flexural) and maturity index at regular
strength intervals. The temperature history of the in-place
concrete is monitored continuously and the in-place matu­
rity index (temperature-rime factor or equivalent age) is
computed from the recorded temperature history. The
in-place strength can be estimated from the maturity index
and strength-maturity relation hip. There are in trument
that automatically compute the maturity index; however,
becau e the value of T0 or Q used by some in tmments may
not be appropriate for the concrete in the structure, these
should be used with care. A TM
1 074 gives the proce­
dure for using the maturity method and provides examples
to illustrate calculation of the temperature-time factor or
equivalent age from the recorded temperature history of
the concrete. A l 306R illu trates the use of the maturity
method to estimate in-place strength during cold-weather
concreting operations. CJ 325 .9R discusse the use of the
maturity method for e timating when the in-place flexural
strength i ufficient to open a pavement to traffic.
The maturity method is intended for e tirnating trength
development of newly placed concrete. trength estimate
are ba ed on four important assumptions:
( l ) The maturity function consta11ts (datum temperature or
Q-value) accurately reflect the temperature dependence of
strength development.
(2) There is ufficient water for continued reaction of the
cementitiou materials.
( 3 ) The concrete in the stlucture is the same a that used to
de elop the trength-matUJity relationship.
(4) The trength potential of the concrete in the structure
is the same as that used to develop the trength-maturity
relationship.
The accuracy associated with the first asswnprion can be
increased by detennining the maturity constants experimen­
tally in accordance with the procedures in A TM C I 074
�
0
0
('0
L1.
c:
.Q
(/)
�
Q)
>
c:
0
()
Q)
C>
<{
15
7
6
5
4
3
2
1
0
0
10
20
30
Temperature, o c
40
50
Fig. 3. 7h-Age conversion factorfor different Q-va/ues and
specified curing temperature of 23°C based on Eq. (3. 7b).
for the particular concrete. Proper curing procedures ( CI
308 R) will ensure that water is available to sati fy the
econd condition. Careful control of eoncret proportions in
accordance with A TM C94/C94M, and proper placement
and consolidation are needed to satisfy the third assumption;
however, explicit verification of constituent proportions is
not always practical. The trength potential of some concrete
mixtures can be significantly affected by high early-age
temperature ( arino 1 984 2004; Brooks et a!. 2007). Thus,
there could be some deviation in trength potential if the
field concrete is placed and cured initially at temperatures
substantially different than used in developing the strength­
matulity relation hip. The third and fourth a umptions
require additional confirmation that the concrete in the true­
lure ha the correct strength potential. This can b achieved
by pcrfonning early-age trcngth te ts on concrete sampled
from the structure in accordance with STM C9 1 8/C9 1 M
or by pcrfonning other in-place te t that gi c indications of
the actual strength level. Such verification i essential when
c timatc of in-place trcngth arc u ed for timing critical
operation uch a formwork removal or application of po t­
tcn ioning. It may also be prudent in omc case to de clop
trcngth-maturity cur c for different initial curing tempera­
ture (Brooks ct al. 2007) or verify lhe curve based on field
infonnation aflcr placcm nt begins {Tex-426-A 20 I 0).
Although the maturity method is most often used to esti­
mate eompres ive strength development, it can be used to
estimate the development of other properric that arc related
to the extent of reaction of ccmcntitious materials with
water. The method ha been u ed to estimate the develop­
ment of concrete ten i le trcngth or flexural trength ( ACI
" 25 .9R- l - ; Bergstrom 1 95 3 · Whiting ct al. 1 994· Hoerner
and Darter 1 999; Delatte ct al. 2000) early-age modulu
of elasticity ( Pinto and Hover L 999a), and overlay bond
trcngth ( Delatte ct al. 2000). Pinto and I lo cr ( ! 999b) and
Wade et al. (20 l 0) demon tratcd the applicability of the
maturity method for the estimation of times of ctting under
different curing temperature . Wade et al. (20 I 0) however
indicated that the apparent activation energy for setting rime
may be lower than for trcngth development.
American Concrete Institute - Copyrighted © Material - www.concrete.org
16
REPORT O N METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
Table 3.9-Useful compressive strength ranges for
In-place test methods
-
I
[Probe
Range of compressive strength*
MPa
Test method
1 0 to 40
Rebound number
penetration
Pin penetration
I
Pullout
ltro oni pul c
velocity
ln summary, the maturity method is used to e timate
strength de elopment during con truction. Becau e thi
method relie on the mea urement of in-place temperature
with re pect to time, and the a umption that the trength­
maturity relation hip is alid, other information i required to
en urc the in-place concrete ba the as umcd trcngth poten­
tial. The correct datum temperature or Q-valuc i required to
impro e the accuracy of the trength e timatiou at early age .
3.8-Cast-in-place cyl i nders (ASTM C873/C873M)
Thi i a technique for obtaining cylindrical concrete
pecimens from oewly ca t slab without drilliog core . The
method i described in A T
7 /C873M and involve
u ing a mold, a illu trated in Fig. 3.8. The outer lee e i
nailed to the formwork and i u ed to upport a cylindrical
mold. The slee e can be adju ted for different lab thickne se
with the top of the lccvc flu h to the floor elevation. The mold
i filled a the lab i ca t, and the concrete ubjectcd to imilar
con olidation and curing operation a the lab concrete. The
concrete in the mold i allowed to cure ith the lab.
The objective of the technique i to obtain a te t pecimen
that ha been ubjected to the arne placement and thermal
hi tory a the concrete in the tructure. To determine the
in-place trength, the mold i removed from the leeve at the
time of te ring and tripped from the concrete cyl inder. The
cylinder i te ted in compre ion in accordance with ASTM
73/ 7 M. For ca e in which the length-diameter ratio
of the cylinder i I . 75 or le , the mea urcd comprc i c
trcngth are corrected by the factor in
TM C42/ 42 M.
Strength r ults of ca t-in-place cylinders ha e b en fi und to
be, on average, 5 to 20 percent higher than tho e of companion
core removed from the general icinity of the cylinders (Popo­
vtc et al. 20 1 6· Bloem 1 96 ). The actual trength difference
\ ill depend on the concrete age material damage caused by the
coring proce , pre ence of embedded reinforcing bars within
the cores, and the manner in which the core are treated during
the time between e traction and te ting (A I 2 1 4.4R ).
In ummary, because the ca t-in-place cylinder technique
involve a comprc ive trcngth te t of a cylindrical pec­
imen, a trcngtb relation hip i not required. To obtain an
accurate e timate of the in-place trength, are i required to
-
f1-
10 to 1 20
3 to 30
2 to 1 30 1
1 500 to 6000
-
-
I to 70
I 00 to 1 7,000'
500 to 4000
300 to 1 9,0001
1 00 to 1 0,000
o limit (docs not mcm.urc compressive
Pull-ofT
Fig. 3. 8----special mold and support hardware to obtain cast­
in-place concrete specimen.
psi
strength)
Maturity
o limit
Cast-in-place cylinder
No limit
. .
Htghcr strengths may be tested 1f satisfactory datu arc presented for the test method
and equipment to be used.
'For •Lrcngth abo\c 40 MPa (6000 p i). pcctul probes arc rcqutrcd.
'for >trengths above
extra 1 pullout inscn
55 MPa ( 000 psi).
spe ial lugh- trength bolt£ are required to
ensure that the concrete in the mold i properly consol idated
in accordance with A TM 873/ 873 M .
3.9-Strength limitations
Most test procedure have orne l imitation regarding the
applicable trength range. I n some ca e , the te t apparatus
has not been designed for testing low-strength or high­
strength concrete, and in other cases there is l imited experi­
ence in u ing the methods to test high-strength concrete. The
u eful strength range for the arious method are umma­
rized in Table 3 .9. The e range are approximate and can be
extended if the user can show a reliable trength relation hip
at higher trength .
3.10-Combined methods
The term "combined method" refers to the u e of two or
more in-place te t method to e timate concrete trength. By
combining results from more than one in-place test, a multi­
ariable correlation can be established to estimate strength.
ombined methods are reported to increa e the reliability
of the e timated strength. The underlying concept is that
if the two method are influenced in different way by the
same factor, their combined use results in a canceling effect
that improves the accuracy of the estimated trength; for
example, an increa e in moisture content increa es pul e
elocity but decrea es the rebound number.
ombined methods were developed and have been used
in Ea tern Europe to evaluate concrete strength in existing
construction or in precast elements (Facaoaru 1 970, 1 984;
Teodoru 1 9 6, 1 9 8). ombination of test methods, such as
pulse elocity and rebound nwnber (or pulse velocity, rebound
number, and pulse attenuation) have resulted in strength rela­
tion hips with higher correlation coefficients than when these
methods are u ed individually. The improvements, however,
have u ually only been marginal (Tanigawa ct al. 1 984; amarin
and Dhir 1 9 4· amarin and Meynink 1 9 I ; Teodoru 1 988).
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
Breysse ( 2 0 1 2 ) conducted a comprehensive critical
review of available data and regression models for the
combined method of rebound hammer and ultrasonic pulse
elocity. Brey e noted that the quality of trength e tima­
tion from any in-place test depend on three element : I )
the sensitivity of concrete strength to the property measured
by the in-place test; 2) the range of the in-place te t values
and corre ponding concrete strength alue used to estab­
li h the trength relation hip; and 3) the mea urement etTor
a ociated with the in-place te t which may overwhelm
errors associated with the regression model. From the litera­
ture review, Breysse concluded that attempts to identify a
univer al trength relation hip are simply wasted energy.
For a given et of data everal regression models will lead
to similar strength e timate , but the key is to adapt the coef­
ficients of the model to provide the be t estimates of actual
in-place trength as oppo ed to trength estimates of labora­
tory specimens. Brcysse explained why combined methods
do not alway re ult in an improvement in the quality of esti­
mated trength : to be beneficial, each of the method u ed
in combination have to provide the same quality of infonna­
tion about concrete strength.
Another approach is to use the maturity method in combina­
tion with another in-place test that measures an actual trength
property of the concrete, uch a a pullout test. The maturity
method i u ed to determine when the in-place concrete should
have reached the required trength, then the other te t method
i cariied out to verify that the strength has been achieved.
This approach is especially beneficial when in-place te ts
involve embedded hardware. The use of the maturity method
to determin � hen the other te t hould be performed may
avoid premature te ting. In addition, maturity readings can
be u ed to as es the significance of lower or higher than
expected in-place test results ( outsos et al. 2000).
ote that combining methods i not an end in itself; a
combined method should be used in those ca es where it
i the mo t economical way to obtain a reliable e timate of
concrete strength ( Le hchin ky 1 99 1 ). In North America,
the use of combined methods ha hown minimal intere t
among researchers and practitioners. There have been no
efforts to develop ASTM tandard for their use.
3.1 1 -Summary
Methods that can be used to estimate the in-place trength
of concrete or to measure bond strength of repair materials
and overlay have been reviewed. While other procedures
have been proposed ( Malhotra 1 976; Bungey et al. 2006;
Malhotra and Carino 2004), the discus ion has been l imited
to tho e techniques that ha e been tandardized a A TM
test method .
Table 3 . 1 1 summarizes the relative performance of the
in-place te t di cu ed in thi report in term of accuracy
of e timated trength and ea e of u e. The table al o indi­
cates which methods are applicable to new con truction
and which are applicable to existing constmction. Gener­
ally tho e methods requiring embedment of hardware are
l imited to use in new construction. In general those tech­
nique that involve preplanning of test locations and embed-
17
Table 3.11 -Relative performance of in-place tests
I
Ease of
A TM
Test method
use'
Rebound number
1--
Penetration
H
resistance
Pullout
900
H
Pulse velocity
597
H
H
+
+
Ht
H
test method with a-++ results in a more accurate strength estimate or is easier to usc than
a method with a +.
I
indicates that the method 1
'Rcqutrc. vcrincation by other tests.
not appltcable to e'istmg constnuctton.
ment of hardware require more effort to u c. Tho e method ,
hO\ e er al o tend to gi e more reliable trength e tirnate .
The u er hould consider the relative importance of accuracy
and ea e of u e hen electing the mo 1 appropriate in-place
te ting y tem for a particular application.
In-place test provide alternative to core test for e ti­
mating the trength of concrete in a structure or can upple­
ment the data obtained from a limited number ofcore . The e
methods are ba ed on measuring a concrete property that has
orne relation hip to trength. The accuracy of the e method
i , in part, determined by the degree of COITelation between
trength and the pby ical quantity mea ured by the in-place
te t. For proper evaluation of te t re ult , the u er bould
be aware of tho e factor other than oncrete trengtb that
n affect the te t re ult . dditional fundamental re earch i
needed to improve the under tanding of bow the e method
are related to concrete strength and bow the test results are
affected by factors other than trength.
An e entiat tep for using the e method to e timate
the in-place trength i the development of a relationship
between trength and the quantity mea ured by the in-place
te t. The data acquired for de eloping the trength relationhip pro ide aluable infom1ation on the reliability of the
e timate . Sub cquent chapter of thi report di cu
the
tali tical cbaracteri tic of the te ts, method for de eloping
trength relation hip , planning of in-place te t , and inter­
pretation of the re ult . The final chapter deal with the u e
of in-place te ts for acceptance of concrete.
CHAPTER 4-STATISTICAL CHARACTERISTICS
OF TEST RESULTS
4.1 -Need for statistical analysis
In de igning a tructure to afely re i t the e pected
load , the engineer usc the pecified compre i e trengtb
f of the concrete. The trength of the concrete in a truc­
t ure i
ariable and, a indicated in AC I 2 1 4R, the peci­
fied compre ive trength i e peeted to be e ceeded with
.
c'
American Concrete Institute - Copyrighted © Material - www.concrete.org
18
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228. 1 R- 1 9)
about 90 percent probability· that is, no more than about I 0
percent of strength test re ults are expected to be le than
the pecified strength. Thus the specified strength can be
considered a the tenth-percentile trength. To ensure that
no more than I 0 percent of the te t results are below f/, the
concrete supplied for the structure hould have an average
standard-cured cyl inder strength that i greater than fc' as
di cus ed in A I 2 1 4R. The difference in required concrete
cylinder trength and de ign compre ive trengtb depend
on the variability of the producer's previou trength-te t
results for similar concrete.
In assessing the abi lity of a partially completed structure
to re i t con truction load , the committee bel ieves that the
tenth-percentile in-place compre sive trength (strength
exceeded with 90 percent probabil ity) should be used for
comparison against the required compressive strength at the
time of application of the construction load . The required
strength mean the compre sive strength used in computing
the nominal load re i tance of structural element at the
applicable load tage.
The use of the tenth-percentile value to interpret re ult
from in-place tests in partially completed structures is con id­
ered reasonable by u ers of in-place tests. The critical nature
of construction operations in partially completed structures
the ensitivity of early-age strength on the previous thermal
history of the concrete, and the general lack of careful
con ideration of construction loading during the design of
a tructure dictate the use of a con ervative procedure for
e aluating in-place test results. For ituations where the
con equence of a failure may not be serious, the e timated
mean trength may be an a ceptable mea ure to as e
the adequacy of the in-place strength for proceeding with
con !ruction operations. Examples of such situations would
include slabs-on-ground, pavements, and some repairs.
I nadequate trength at the time of a proposed construction
operation can u ually be remedied by simply providing for
additional curing before proceeding with the operation.
In-place test may al o be u ed to evaluate the concrete
strength in a structure when que tion ari e because of low
strengths of standard-cured cyl inders during construction. I f
the compre i e trength of standard-cured cylinder fails
to meet ACI 3 1 8 acceptance criteria, te ting of core may
be required. Ba ed on C I 3 1 if the average compres ive
strength of three cores taken from the questionable portion
of the structure exceeds 85 percent of the specified compres­
sive strength and no single core strength is less than 75
percent of the specified trength, the concrete strength is
deemed to be acceptable. There are, however no analogous
acceptance criteria for the e timated in-place compre i e
trength ba ed on in-place test . Chapter 8 di cu se how
in-place testing could be used for acceptance of concrete.
l rre pecti e of the objective of in-place te ting, a reli­
able e timate of the in-place compre ive trength require
the use of tati tical methods that account for the following
primary ources of unce11ainty:
(a) The a erage value of the in-place test re ult
(b) The relationship between compressive strength and the
in-place test re ults
(c) The inherent variability of the in-place compressi e
strength
The first source of uncertainty is associated with the
inherent ariability (repeatabil ity) of the te t method which
i discus ed in this chapter. haptcrs 5 and 7 addres the
other main source of uncertainty.
4.2-Repeatability of test results
The un ertainty of the average value of the in-place te t
re ult i a function of the tandard de iation of the re ult
and the number of test . The standard deviation is, in tum, a
function of the repeatability of the test and the ariability of
the concrete in the structure.
In this report, repeatability means the standard deviation
or coefficient of variation of repeated te ts by the same oper­
ator on the ame material. This is often called the ingle­
operator variation and how the inherent catter a ociated
with a particular test method.
Data on the repeatabi lity of some in-place te ts are
provided in the preci ion tatement of the A T
tandard
go erning the test . orne information on the repeatability of
other tests may be found in published reports. nfortw1ately,
mo t published data deal \ ith correlations with standard
strength test , rather than with repeatability. s will be seen,
conclu ions about repeatability are often in conflict becau e
of differences in experiment de igllS or in data analysis.
4.2. 1 Rebound number-The precision statement of
A TM 805/ 805M tate that the ingle-operator tan­
dard deviation of the rebound hammer test is 2.5 rebound
numbers which limit the expected range of 1 0 readings
to 1 2. The re ult of three studie that evaluated the perfor­
mance of variou in-place te t provide additional in ight
into the repeatabi l ity of the rebound number te t. Keiller
( 1 9 2) used eight different mixtures and took 1 2 replicate
rebound reading at age of 7 and 2 day . Carette and
Malhotra ( 1 984) used four mixture and took 20 replicate
readings at ages of I , 2, and 3 days. Yun et al. ( 1 9 ) u ed
five mixtures of concrete and took 1 5 replicate readings at
age ranging from I to 9 1 days.
F igure 4.2. 1 a shows the tandard deviations oft he rebound
number as a function of the average rebound numb r. The
data from the three tudie appear to follow the arne pattern.
In the study by Carette and Malhotra ( 1 984), the average
maximum rebound number ranged from 1 5 to 22 and the
average standard deviation was 2.4. In the study by Keil ler
( 1 9 2) the average rebound number ranged from I to 35,
and the average standard deviation was 3 .4. I n the work by
Yun et al. ( 1 98 ) the range in average rebound number was
1 2 to 32 and the average tandard deviation wa 2.5.
Examination of Figure 4.2. 1 a how that there may be
a trend of increasing standard de iation with increasing
average rebmmd number, in \ hich ca e the coefficient of
ariation i a better mea ure of repeatabi lity. Figure 4.2. 1 b
shows the coefficients of variation plotted a functions of
average rebound number. There doe not appear to be any
trend with increasing rebound number. In contra t, Le h­
chin ky et al. ( 1 990) found that the coefficient of varia­
tion and its variability tended to decrea e with increasing
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228. 1 R-19)
concrete strength. The average coefficients of variation from
the tudies by Carette and Malhotra ( 1 9 4) and by Keiller
( 19 2) have equal values of 1 1 .9 percent while the average
alue from the tudy by Yun et al. ( 1 9 ) wa I 0.4 percent.
6
•
5
c:
0
�
·
:;;:
Q)
0
"0
.....
(II
"0
c:
(II
U5
•
Y un et a l
.
( 1 988)
.
�
4
....... ·-········
3
·········· ··········· ··········
:�
2
�
· · · · · ·
10
· · ··
.
: ·
; r
· ··
'
.
r r�
..
· · ·
··
· · · · · · · · · · · · · · -:- · · · • • • • • • •
:·
·
�
· =
�
·
'·· ·······
.
·· · · · · · · · ·.· ·:-: · · · · · · · · · � · · · · · · · · · ·
:. .
:. . .. .,.
. • .. . ,. «:
.
:
�: ·
��
·
:
�
�
.
:�
.
.
�
·· · · ·
:
.
(1 982)
Carette & Malhotra (1lj84)
Ketller
•
· · ·
15
· · ··
-
-
-- -
· ·
25
20
T·
· · · · · · ·
T
· · · · · · · · ·
35
30
40
Average Rebound Number
Fig. 4. 2. /a-Sing/e-operator standard deviation as a func­
:
tion ofaverage rebound numbe1:
20
�
::
··�;::
�
0
c:
Q)
·u
=
Q)
0
(.)
.
.
,----o----�-=--�---,
15 +
·······
.
.
+
10 +··········
5
+
:
···
· ·
· · · · · ··
····
.
··········
.
:1
· · · · · ·· ··· ········· ··
•
�
.·
•••
�
L....... .
..:
· · · · · · · · · · - - - - - - - · · · · · · · · -•
98M-�-Iho-tra__:(�
...----=-'---�e-�-=-:...:.�1 984
--,)
-
:
=· +
+ ·
:
! • • . ·,·
t I
:
�
� � .....
�
:� :. ... �:
... .
� .
: ..
. : . . :·
··· · ·· ·· ·
. .....
;
·
r . T ......... .
et
(1988)
•
o �����.��+=��i��+���
Yun
10
15
al
20
25
I
30
.
40
35
Average Rebound Number
:
Fig. 4.2. 1 b--Single-operafor coefficient of variation as a
function ofaverage rebound numbe1:
In Figure 4.2 . 1 b the coefficients of variation are not
con tant. ote, however, that the values are based on sample
estimate of the true averages and tandard de iation . With
finite ample ize there will be variation in the e e timate ,
and a random variation in the computed coefficient of varia­
tion i e pected. although the true coefficient of ariation
can be con tant. Thus, it appears that the repeatability of the
rebound number technique can be de cribed by a constant
coefficient of variation, which ha an a erage value of
approximately I 0 percent.
4.2.2 Penetration re istance-The single-operator preciion statement in A TM
03/C 03M for probe and pin
testing are urnmarized in Table 4.2.2.a and Table 4.2.2b,
re pecti ely.
The data reported by arette and Malhotra ( 1 9 4) and
Keiller ( 1 9 2), which include concrete strength in the range
of I 0 to 50 M Pa ( 1 500 to 7000 p i), give additional in ight
into the underlying mea ure of repeatability for this te t.
Figure 4.2.2a how the tandard deviation of the expo ed
length of the probe a a fun tion of the average expo ed
length. The values from arette and Malhotra ( 1 9 4) are
based on the average of six probe , while Keiller ( 1 9 2 )
r e ults are based o n three probes. xcept for one outlying
point, there i a trend of decreasing single-operator vari­
ability with increasing exposed length. In Fig. 4.2.2b the
coefficients of variation of exposed length are shown a a
function of the a erage exposed length. The decrea ing trend
with increa ing concrete trength is more pronounced than
in Fig. 4.2.2a. Thus, the repeatability of the expo ed length
i described neither by a constant standard deviation nor a
con tant coefficient of variation.
The cu tomary practice i to measure the expo ed length
of the probes but concrete trength has a direct effect on the
depth of penetration. A more logical approach is to express
the coefficient of variation in terms of depth of penetration.
Figure 4.2.2c show the coefficient of variation of the penetra­
tion depth a a function of average penetration. In thi case,
there is no clear trend with increasing penetration. The higher
catter of the values from the Keiller ( 19 2) te t might be due
to their smaller sample size compared with the tests of arette
and Malhotra ( 1 984). ote that the standard de iation ha
the arne value whether expo ed length or penetration depth
i u ed. The coefficient of variation, howe er, dep nd on
Table 4.2.2a-Single-operator precision for penetration-resistance testing with probes (ASTM C803/C803M)
I
I
l
I
I
Maximum expected difference bel\ cen two tests (each
calculated as the average of three measurements)
Maximum size of aggregate
Standard deviation
Mnximum range of three
individual measurements
Mortar--4.75 mm ( o. 4 )
2.0 m m (0.08 in.)
6.6 111m (0.26 in.)
3.3 mm (0. 1 3 in.)
oncrete-25 mm ( I in.)
2.5 111111 (0. 1 0 in.)
.4 111111 (0.33 in.)
4.1 111111 (0. 1 6 in.)
50 mm (2 in.)
3.6 mm (0. 1 4 in.)
1 1 .7 mm (0.46 in. )
5.6 mm (0.22 in.)
Concrete
Table 4.2.2b-Single-operator precision for penetration-resistance testing with pins (ASTM C803/C803M)
Maximum range of six individual
Material
Con ret
3 to 28 MPa
(450 [0 4000 p i)
tandard deviation
0.4 mm (0.0 1 6 in.)
measurement
1 .6
nun
19
(0.064 in. )
Ma imum expected differcn e bet\\CCn two tests (ca h
calculated a the average of ix measurement�)
0.5 mm (0.0 1
American Concrete Institute - Copyrighted © Material - www.concrete.org
in.)
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
20
Average Exposed Length, in.
04
08
12
16
2
24
28
5.0
+-.,....,...
.,.... ....
.f ,...
.. +-.
..,.. ,....-f
...,.. -.-....+-.
..- ,....-f
...,.. -.-....+
..- 0 20
40
0 16
�
3.0
0 12
'E
2.0
0 08
1 0
0 04
E
E
c
0
!:
c
.Q
iii
;;
�
·;;
�
01
"0
c
g
(/)
40
30
20
Average Exposed
Length ,
01
"0
c
g
(/)
0 00
60
50
'E
10
mm
Fig. 4. 2. 2a-Single-operator standard deviation a a .fimc­
tion ofaverage e.xpo ed length ofprobes.
Average Exposed Length, in .
0.8
1 .2
1 .6
2
2.4
2.8
0.4
.- ���
14 1-�-r�rr,-���-r,-�....�
:
Q)
·o
�
8
!
:
T� J�
�� :: .
�
!
I: I r •••
.. . .
: • . . T• .• .• i � ?r. · · · · r· · · · r • •
· ·
. . "t "�'""�'""""'
��
,
4
•
2
•
:
:
:
: •
-�. ·
.
.
:
• .;
..
.:
:
••
:
• .
.......l . . � ........
; ... . ... . .. .
•
Kcdler (1982)
CareHe & Malhotra (1984)
•;
:
•
i���
i ���
: ��
i���
o ������
10
40
30
20
50
60
70
Average Exposed Length, mm
Fig. 4.2.
ingle-operator oef]i ienl of varia/ion a a
function ofaverage exposed length ofprobe .
Average Penetration, in.
0.8
1 .2
0.4
1 5 4-rT,-�,-�......
1 .6
c
0
-
�
·c
ro
.
i
10
>
0
c
Q)
·o
:e
• •
.
.. ..... � ..........! ....
•
. ':
.
.
. .
-
: .
Q)
0
(.)
•
10
20
!•
�
30
40
:
..
.
: --- . - · · - -·-�- - . -...... -�---------.
f
i
.
2.8
Ka.ller (11182)
•
.
2.4
....�
.. ....�
.. -r��
�
•
..... :f . . . . . . . -f- ...... . .
5
2
�
�
-.
:•
.
�.
l
.
:
l
. . .. . . . . . . ill
.
:
�
. . � .........
:.
.
i
;
50
60
70
Average Penetration, mm
Fig. 4.2.2c ingle-operator coefficient of varia/ion as a
June/ion ofaverage pene/ration ofprobes.
whether the standard deviation is divided by average exposed
length or average penetration depth.
Hence, it appears that a con tant coefficient of variation
of the penetration depth can be u ed to de cribe the ingle­
operator variability of the probe penetration te t. The work
by Carette and Malhotra ( 1 9 4) is the first known study
that u es tbjs method for defining the repeatabil ity of the
penetration te t. Other te t data u ing the probe penetration
ystem, ho ever, can be manipulated to yield the coefficient
of ariation of penetration depth, provided tv o of the e three
quantities are given: a erage exposed length, standard devi­
ation, or coefficient of variation of exposed length. sing
the data given in Table 6 from the Malhotra ( 1 976) review
the following alues for average coefficients of variation for
depth of penetration are calculated:
-
Maximum izc aggregate
Coefficient of variation of penetration
mm
in.
depth, percent
50
2
14
25
I
.6
19
314
3.5, 4.7, and 5.6
l n the rudy by Carcttc and Malhotra ( 1 9 4), the maximum
aggregate size wa 1 9 mm (3/4 in.) and the a cragc coeffi­
cient of ariation wa 5 .4 percent. \ hcrcas in the study by
Keillcr ( 1 9 2 ), it was 7.8 percent for the arne maximum­
size aggregate. Other work (Swamy and Al-l lamad 1 984)
used I 0 rnm (3/ in. ) maximum- izc aggregate, and the coef­
ficient of variation ranged bct\vecn 2.7 and 7 percent. For
commonly used 1 9 mm (3/4 in.) aggregate, it is concluded
that a coefficient of ariation of 5 percent is rca onablc.
There arc limited data on the repeatability of the pin pene­
tration te t. a er and !-Mana ecr ( 1 9 7b) reported an
average coefficient of variation of approximately 5 percent
for replicate tc t on 1 00 mm (4 in.) thick lab pecimen
and on the bottom urfaces of 1 50 x 300 mm (6 x 1 2 in.)
cylinders. The variability wa ba ed on the b t five of seven
readings with the lowe t and rughc t d leted; the concrete
strength aried from approximately 3 . 5 to 25 M Pa (500 to
3500 p i). I n another study (Carino and Tan k 1 9 9) eight
replicate pin tc ts were petfomlcd at the midhcight of I 00
x 200 mm (4 x 8 in. ) cylinder . The comprc sivc strength
ranged from approximately 7 to 40 M Pa ( I 000 to 5800 psi).
ach et of replicate pin tc t wa analyzed for outlier due
to penetration into large air void or coar c aggregate parti­
cle . On average, two of the eight readings were discarded.
Figure 4.2.2d shows the standard deviation ofvaHd penetra­
tion value plotted a a function of a crage penetration. ote
that a rugh pen tration eorrc pond to low concrete trcngtb.
There is no clear trend between the standard deviation and
average penetration. The average tandard deviation is 0.4 1
mm (0.0 1 6 in.) which i the value adopted in the precision
statement of A TM 803/C803 M . To compare with the
atiability reported by a cr and Al-Mana ecr ( 1 987b), the
result in Fig. 4.2.2d arc pre entcd in term of coefficient of
ariation in Fig. 4.2.2c. The average coefficient of variation
i 7.4 percent. Ba ed on available infom1ation, a coefficient
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
Average Penetration , in.
0. 1 4
0.16
.
0. 1 8
1 00
E
E 0.75
c
.Q
a;
·:;; 0 50
Q)
0
-e
m
"
t:
!!!
:
0.2
0.22
t r- r ...
�-r .J.... .....
T
.
.
···
·······
····
···
;;
,_ .
�
. .....
•
.
�
.
�
... .
•
·
.: •
�
0.26
0.28
·
.;
+
65
::��
--
.
r
-
. . .
.
··
·�
!\
.
: . • :
en
3.5
;
;
4.0
4.5
.
�
·
: ..
.... _
:
:
l·
;
5.0
..
• :
•
�
�
5.5
6.0
c
5
·:;;
Cl>
0
"
m
4
en
en
t:
m
.
�
6.5
�
6
c
.2
a;
·:;;
0.02 Cll
0
-e
m
"
0.01
.
.
.s
0.03
....� . . v� ..�.. ·-� j. ....• t. /-:--i •�- . • �j ......
0.25
•
7
WIC • 0 65
•
••
0
Q�
WIC • 0 50
: . . . . . . . ., ..
..
.
Pullout Load, lb
0.24
e
0
7ii
"0
t:
<0
3
2
2000
- : � r r::r· +
� �
. . . . . -r· �� � �:� :: � -.r. :::::: :::.1:::::: :
•
···········-
!
LW
···
�
i
·············
·-···········
�
..
:.!
-
�
(ij
-
>
0
c
Cll
'(j
�
Cll
0
.
.
..
.
...
10
········
:
-
��
....
(.)
.
.
•
. . .• . . . .
.
.
�
:
- - - · · · · · ··
:
�
•
•
- -
WIC = O SO
WIC = 0 55
WIC = 0 60
•
0
10
.
.
.. ,
..
•
.
-
4.0
4.5
5.0
5. 5
·-··
��
�
· :
--·
·
······
-·
-
.
�
.
·
.
·
·:;;
Cll
0
-e
Ill
"
675
· - 450
t:
<0
en
225
.
30
0
�
900
..
:9.
c
50
40
.
.
..
-
..
.
6.0
6. 5
7.0
Average Penetration, mm
�
. . . . .. . .
- - - - ·
.
..........
_
: �
;
�
. . . . . . . . . . . . • . . . . . . . . . . . .i
....
.
:
!: • • A ;: • .
.
.
. .
. .
Q)
.
1
6000
.
8000
..
.�
..
......
:
:
:
.
.
.
.
.
• . . . . . . . . . . . ; . . . .......... � . . . . . . A. . . . . .
0
c
.
4000
:
�.
1 0000
... . . . .
� ..........
.
..
:
.
:
. . . . . . . . . . . ., . . . . . . . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . . . . . � . . . . . . . . . . . . .
·;;::
(tJ
,
2000
..
:
c
o �����i�����i���i=��
3.5
·
20
............�
>
•
�
--
;
···········
1 1 25
4
: ��-r��-r 0
i�
• �
i ��•,_
o ���,_
,_
.2
a;
-- Average
. :
.
�
·
t
1 350
Pullout Load , lb
I
• we
:
:
: • :
J . ..;.: .,. ...I'J .:. .•. . . . . . : . . . . . ...: ..• . .. .•:: . . . ..
. . .: . ..
:
:. . • : . 1 :
.:
:
• -�
l
!_
:
-.�·
.
:•
.
e
5
.;.........1
•
1 575
Fig. 4. 2.3a-Single-operator standard deviation as a func­
tion ofpullout load (Stone et a/. 1 986).
0.14 0. 1 6 0.1 8
0.2
0.22 0.24 0.26 0.28
.
�
20 �rr�
�
rr
+
�
�
rT
�
!:
!-rrT�-rr+��
!
!.
15
10000
·
Average Penetration, in.
c
.2
8000
Pullout Load, kN
Fig. 4.2. 2d-Standard deviation ofpin penetration test on
100 x 200 m m (4 x 8 in.) cylinders (Carillo and Tank 1 989).
0
6000
������
0
7.0
4000
Average Penetration, mm
�
21
· · · ·· · · ·· · · ·
· · · · · · · ···
·
:
�
• :�
.. . .
....
: ....
· · · · · · · -
· · · · ·· ·· · · · · ·
- -
...
............ ]
.
!
0
10
.� .
.
. . ......
.
.
...
.
20
...
..
.
....
..
..
•
.
.
""'
�
..
.
:
!
30
•
•
:.I
•
�
�
-- ·
· · · · · · · ·
'to' ' � • • • • • • • • • • • • •
:
· · · · · · · ·· · · · ·
t . . .i
............ j ............. .:. . . . .
'(j
!E
Q)
0
(.)
.
:
;
. . . . ... . . . . . . . .
G70
G54
L$
...
. t. .
LW
40
50
Pullout Load , kN
Fig. 4.2. 2e-Coefficient of variation ofpin penetration tests
on 100 x 200 mm (4 x 8 in.) cylinders (Carillo and Tank 1 989).
Fig. 4. 2. 3b---Single-operator coe.ficienl
f
of variation as a
function ofpullout load (Stone et a/. 1 986).
of variation of 8 percent is recommended for planning pin
penetration tests.
4.2.3 Pullout test-1\.STM C900 tates that the average
single-operator coefficient of variation is 8 percent for cast­
in-place pullout tests with emb dments of about 25 mm ( J
in.) in concrete with nominal maximum aggregate izc of 1 9
mm (3/4 in.). Thi value i based on the data summarized a
follows. A similar single-operator variabi lity is suggested for
post-in tailed test of the same geometry (Peter en 1 997 ).
tone ct al. ( 1 986) examined whether tandard deviation
or coefficient of variation i the best mea urc of repeat­
abil ity. Four tc t series were pcrfonncd. Three of them used
a 70-dcgrcc apex angle (2a) with different aggregate type :
sil iceous river gravel, crushed limestone, and expanded
lm: -den ity (lightweight) shale. The fourth cries was for
a 54-degree angle with river-gravel aggregate. These test
series arc identified as G70, LS, LW, and G54 in Fig. 4.2.3a
and 4.2.3b. The embedment depth was about 25 mm ( I in.)
and compressive strength of concrete ranged from about I 0
to 40 M Pa ( 1 500 to 6000 psi). Figure 4.2.3a show the tan­
dard deviation, using 1 1 replication , as a function of the
average pullout load. ote that there is a tendency for the
standard deviation to increase with incrca ing pul lout load.
Figure 4.2.3b hows the coefficient of variation as a function
of the a erage pullout load. I n thi ca e, there i no trend
between the two quantities. Thus, it can be concluded that
the coefficient of variation should be used a a mea urc of
the repeatabi lity of the pullout test.
Table 4.2.3a gives the reported coefficients of variation
from different laboratory studies of the pullout te t. Be ides
these data, the work of Krenchel and Petersen ( 1 984)
ummarizcs the repeatability obtained in 24 correlation
testing programs involving an insert with a 25 mm ( I in.)
embedment and a 62-dcgrec apex angle. The reported co f­
ficients of ariation ranged from 4. 1 to 1 5 .2 percent, with an
average of 8 percent. The tests reported in Table 4.2.3a and
American Concrete Institute - Copyrighted © Material - www.concrete.org
<aCi'J
22
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
Table 4.2.3a-Summary of single-operator coefficient of variation of pul lout test
-
--
Embedment depth
pex angle,
dcg
Reference
Malhotra and
67
Carcnc ( 1 980)
Malhotra ( 1 975)
67
-
Bickley ( 1 982b)
Khoo ( 1 9 4 )
arcrtc and
Malhotra ( 19 4)
Keiller ( 1 982)
62
r--
70
67
tone et al. ( 1 9 6)
Ls
occa ( 1 9 4}
in.
0
2
50
2
25
I
mm
in.
25
I
I
6
type
10
Limestone
19
50
2
19
3/4
Granite
0.9 to 1 4.3
5.3
2.3 to 6.3
-
-
-
3.9
-
4. 1
3/4
Lime tone
-
-
314
Limestone
62
25
I
19
3/4
Limestone
1 .9 to 1 2.3
6.9
4
1 .9 to I I .
7. 1
10
5.2 t o 1 4.9
8.5
6
7.4 to 3 1
1 4.8
6
-
19
-
-
I
19
II
4.6 to 1 4.4
1 0.2
25
I
19
34
Limestone
II
6.3 to 1 4.6
9.2
19
314
Low density
II
1 .4 to 8.2
6.0
19
3/4
Gravel
II
4.3 to 1 5.9
1 0.0
nknown
24
2. to 6. 1
4.3
70
25
I
54
25
I
67
30
1 .2
I
I
L
Gravel
3/4
--
-
--
--
-
13
l /2
--
-
-
25
-
-
Average
3.2 to 5.3
I
-
3
Range
--
25
70
2
-
·-
Coefficient of variation, percent
o. of replicate
specimens
Unknown
/8
-
I
I
ra el
1/4
-
25
Aggregate
62
70
r-
-
mm
Maximum aggregate size
-
-
--
-
-
Table 4.2.3b-Summary of results from i nvestigation of pul lout test (Stone and Giza 1 985)
Test series
Embedment depth
Apex angle.
deg
mm
m.
mm
m.
Aggregate type
25
0.98
19
314
Gravel
2
46
25
0.98
19
314
Gravel
4
54
25
0.9
1-
19
314
Gravel
2
r-
19
3/4
19
314
1-
19
314
19
3'4
Gravel
-
-
2
0.98
25
0.9
25
0.9
25
0.98
58
12
0.47
-
62
-
70
6
-
Embedment
1--
ggregate
size
type
--
Gravel
i-
-
2
2
1 0.3
Gravel
X
11
5.6 to 1 8.7
1 1.1
II
6.3 to 6.7
11
8.6 to 1 0.0
X
X
X
11
1
9.9
-
1 2.9
11
7.7 to 1 4.0
1 0.9
X
11
6.5 to 6.7
6.6
8.8 to 1 0.7
9.8
X
19
3/4
Gravel
2
3/4
Gravel
2
58
25
0.98
19
3/4
Gravel
2x II
X
58
27
1 .06
19
314
Gravel
2
58
43
1 .69
19
314
Gravel
2
70
25
0.98
6
1 14
Gravel
- X
70
25
0.98
10
31
0.9
13
1 2
19
3/4
19
34
70
70
-
-
70
I
25
25
0.9
25
0.98
11-
1-
-
Gravel
--
Gravel
Low dcnsiry
25
0.98
19
314
Gravel
25
0.98
19
3/4
Crushed gnei>"
25
0.98
19
3/4
Porous limestone
-
11
9. 1 to 1 1 . 1
1 0. 1
11
1 1 .5 to 1 1 .9
1 1 .7
11
6.5 to 7.0
6.8
5 x 11
4.9 to 6.5
6.0
3 . 3 to 1 0.6
6.7
5
4
X
X
X
II
11
-
-
8.0 to 1 0 . 1
-
5.6 to 5.7
4x II
8.0 to 1 0 . 1
8.8
X
11
7.2 to 1 6.8
1 2.0
2x II
7.7 to 1 0.9
9.3
2
-
-
-
-
-
-
-
2x II
I I"" indicates two groups of I I replicate per group.
by Krenchel and Peter en in ol ed different test geometrie
and different type and ize of coar e aggregate. In addition
the geometry of the pecimen containing the embedded
in ert wa diffe rent, with cylinders, cube beam , and lab
being ommon hape . Becau e of the e te ting difference ,
it i difficult to draw firm conclu ion about the repeatability
of the pullout te t.
8.6
9.0 to 1 0.8
19
--
-
-
II
0.78
--
7.5 to 9.6
9.
X
0.9 1
1--
-
2x II
23
ravel
-
.0 to 1 0. 1
-
6.5
--
11
4
ravel
-
Average
20
70
'
Gravel
-
Range
58
70
•The tcnn. '"2
f-
-
r-
X
oeffictent of vanation, percent
9. 1 to 1 1 .4
58
70
ggrcgatc
r-
314
X
1-
II
specimens•
-
-
19
-
1-
II
No. of replicate
30
pex allgle
r-
Maximum aggregate size
5.7
-
Table 4.2.3b ummarize the coefficient of variation
obtained in a tudy by tone and Giza ( 1 985) de igned to
e amine the effect of different variable on te t repeat­
abi l ity. The column labeled ample ize how the number
of group of te t , \! ith each group containing I I repl ica­
tion . For the condition tudied, il wa found that embed­
ment depth and apex angle did not greatly affect repeat­
abi l ity. The maximum nominal aggregate ize, however,
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
appeared to have some affect, with the 1 9 mm (3/4 in.)
aggregate resulting in slightly greater variability than the
smaller aggregates. The aggregate type also appears to be
important. For test with low-den ity aggregate, the vari­
abil i ty was lower than for tests " ith normal-density aggre­
gate . In thi study, companion mortar pecimens were also
te ted, and the coefficients of variation varied between 2.8
and I 0.6 percent, with an average value of 6.2 percent. Thu ,
the repeatability with low-den ity aggregate i imilar to that
obtained with mortar.
xperimental evidence suggests that the ariability of
the pullout test should be affected by the ratio of the mortar
strength to coarse-aggregate strength and by the maximum
aggregate size. A aggregate strength and mortar trength
become similar, repeatability is improved. This explains why
the tests re ult by tone and iza ( 1 9 5) with low-density
aggregate were imilar to te t re ult with plain mortar. Re ults
from Bocca ( 1 984), summarized in Table 4.2.3b, al o lend
upport to this pattern of behavior. Jn thi ca e, high- trength
concrete wa u ed, and the mortar trength approached that
of the coarse aggregate. This condition, and the use of mall
maximum aggregate size, could explain why the coefficients
of ariation were lower than typically obtained with similar
pullout test configurations on lower-strength concrete.
In summary a ariety of test data has been accumulated on
the repeatability of the pullout test . Differences in result are
often due to difference in material and te ting condition . I n
general, it appear that a n average ingle-operator coefficient
of variation of8 percent is typical for pullout tests confotming
with the requirement of ASTM C900 and embedment depths
of about 25 mm ( I in.). The actual value expected in any
particular ituation will be affected primarily by the nature of
the coarse aggregate, a discussed in previous paragraphs.
4.2.4 Pull-off resr-A TM 1 5831 l 58 3 M states that an
inter-laboratory testing program to determine the precision
and bias of the test method has not been completed. ay.­
burd and McDonald ( 1 999) published data that provide an
estimate of the single-operator variability. In the study, three
repl icate test were perfonned for each condition. The mean
alue of the pull-off bond strength ranged from 0.4 1 to 3.44
M Pa (60 to 500 psi). Replicate te t " ith the arne failure
mode were u ed to d !ermine the ingle-op rator tandard
deviation for each test condition. ignificant ariation in
pull-off bond strengths were observed between replicate
specimens in the study. The o erall pooled tandard devia­
tion, for all data sets was found to be 0.29 M Pa (42 psi).
4.2.5 Ultrasonic pulse veloci
I n contrast to the
previous test technique that examine a relatively thin layer
of the concrete in a tructure the pulse-velocity method
(using through transmis ion) examine the entire thick­
nes of concrete between the transducers. Localized differ­
ences in the compo ition of the concrete becau e of inherent
ariability are expected to ha e a negligible effect on the
mea ured travel time of the u ltra onic pul e . Thus the
repeatability of this method i expected to be better than the
pre ious techniques.
Table 4.2.5 reports the single-operator variability of pulse­
velocity measurement obtained by different investigator .
23
Table 4.2.5-Single-operator coefficient of
variation of pu lse-velocity tests
oefficicnt of variation. percen1
Reference
;-
Average
0.5 10 1 .5
1.1
O. l lo 0.8
0.4
Bocca ( 1 9 4)
0.4 10 1 .2
0.7
Yun ct al. ( 1 98 )
0.4 IO 1 . 1
0.6
0.2 10 4.0
1 .9
1 . 1 10 1 .2
1 .2
Kcillcr ( 1 9 2)
arellc and Malholra ( 1 9 4)
L
Range
Lc hchin ky ct nl. ( 1 990)
Phoon c1 al. ( 1 999)
TM C597 state that the repeatability of test results is
within 2 percent for path lengths from 0.3 to 6 m ( I to 20 ft)
through Olllld concrete and for differ nt operator using the
arne in trumcnt or on operator using differ nt in trumcnt .
4.2.6 Maturity method-ll1 the maturity method, the
temperature history of the concrete is recorded and used to
compute a maturity index. Therefore, the repeatability of
the maturity indcxc depends on the in tmmcntation used.
One would expect the repeatability to be better when u ing
an electronic matutity meter than when the maturity index is
computed from temperature reading on a strip-chart recorder.
There arc howe cr, no publi hcd data on repeatability of
maturity measurements u ing different in trumentation. The
pre ision of temperature mea uremcnt by the instrument
i not an important issue, provided that step arc taken to
cnsw·c that the instrument i operating properly b fore it is
used. Temperature probes can be embedded in temperature­
controlled water baths to verify that they arc operating prop­
erly. The maturity index, after a given time in the bath, can be
calculated readi ly and compared with the in trumcnt reading.
Of greater importance than accurate temperature measure­
ment is u ing the datum temperate or Q-valuc that rcprc cnts
the temperature sen itivity of the particular concrete.
4.2.7 Ca t-in-place cylinder- This tc t method involve
the dctcm1ination of the compres i c strength of cylindrical
p cimen cured in the pc ial mold located in the nuc­
turc. The repeatability would be expected to be imilar to
other comprc ion tests on cylinders. Little data have been
published. Bloem ( 1 968) reported a ingle-opcrator coeffi­
cient of variation ranging from 2. 7 to 5.2 percent with an
average of 3. percent for three replicate tc ts at ages from I
to 9 1 days. Data from Carino ct al. ( 1 983b ), in which three
replicate cylinders were tested at age ranging from I to 32
day , how an average coefficient of variation of3.8 percent.
A TM
73/C873M tatcs that the single-operator coefficient of variation is 3 . 5 percent for a range of compre si e
trengtb between I 0 and 40 M Pa ( 1 500 and 6000 p i).
CHAPTER 5-DEVELOPM ENT OF STRENGTH
RELATIONSHIP
5.1 -General
Manufacturers of in-place tc ting equipment typically
provide generalized relation hip in the form of graph or
equation that relate the property measured by the partic-
American Concrete Institute - Copyrighted © Material - www.concrete.org
(clci)
24
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
ular te t device to the compressive strength of standard
concrete pecimens. These relationship , however, often do
not accurately represent the specific concrete being tested.
The e relationships hould not be u ed unles their validity
has been established through correlation testing on concrete
similar to that being inve tigated and with the pecific test
in trument that will be used in the investigation. The general
approach in correlation te ling is to perform replicate
in-place te t and tandard trength test at ariou trengtb
level , and then to u e tati tical procedures to e tabli h the
strength relationship. The details, however, will depend on
whether the in-place tests are to be used in new construction
or in existing structure .
tandard pecimens can be cylinders, cubes, or beam .
The in-place te ts are often correlated with the compre sive
strength of cores becau e core strength is the most estab­
li hed and accepted mea ure of in-place trength. ast-in­
place cyl inders are also useful in determining the in-place
strength of ne� concrete, and their use doe not require a
pree tabli hed correlation. The tati tical technique for
e tabli bing the strength relationship are independent of the
type of standard specimen. The specimen type, however, is
important when interpreting the re ults of in-place tests.
5.2-New construction
5.2 . 1 General- For new construction, the preferred
approach is to e tablish the trength relationship by a
laboratory-te ting program that is performed before u ing
the in-place test method in the field. The testing program
typically in ol e preparing tc t specimens using the same
con rete mixture proportions and material to b u ed in
con truction. At regular interval , mea urement are made
u ing the in-place test technique, and the compressive
strengths of standard specimens are also measured. The
paired data are ubjected to regres ion analysis to determine
the best-fit e timate of the strength relation hip.
For orne techniques it might be po ible to perfonn the
in-place test on standard specimen without damaging them
and the specimens can be sub cquently tested for compres ive
strength. Usually, in-place tests are carried out on separate
specimen , and it i extremely important that the in-place te t
and tandard te ts are performed on p cimen having imilar
con olidation and at the same maturity. Thi can be achieved
by u ing curing conditions that ensure similar internal temper­
ature histories. Alternatively internal temperature can be
recorded and te t ages adjusted so that the in-place and stan­
dard te ts are perforn1ed at the same maturity index.
In developing the test plan to obtain a rel iable strength rela­
tion hip, the user hould con ider the fol lowing question :
(a) l low many trength levels (test points) are needed?
(b) Hm many replicate tests should be perforn1ed at each
trength level?
(c) l lo\ hould the data be analyzed?
5.2.2 Number ofstrength level -The nwnber of trength
levels required to develop the strength relation hip depends
on the desired level of preci ion and the cost of additional
tests. ection A. I in ppendix discus es how the number
ofte t point used to develop the strength relationship affects
the uncertainty of the estimated strength. From that discu sion in . I , it was concluded that in planning the correlation­
testing program, ix to nine strength levels hould be consid­
ered. sing fewer than ix strength levels may re ult in high
uncertaintie in the estimated strength and using more than
nine level may not be ju tifiable economically.
The range of strengths u ed to e tabli h the correlation
hould cover the range of strengths that are to be e timated
in the tructure. Thi will en ure that the trength relationhip will not be u ed for extrapolation beyond the range
of the correlation data. Therefore if low in-place strengths
are to be estimated, such as during l ipforming, the testing
program mu t include the e low strength levels. The cho en
trength level
hould be evenly distributed within the
trength range.
5.2.3 umber of replications- The number of replicate
te t at each trength le el affects the uncertainty of the
average value . The tandard deviation of the computed
average aries with the in er e of the square root of the
number of replicate te t u ed to obtain the average. The
effect of the nwnber of te t on the preci ion of the average
is simi lar to that shown in Fig. . I ( ppendix A).
Statistics show (A TM E 1 21) that the required number
of replicate tests depends on: I ) the single-operator vari­
abil ity of the method· 2) the al lowable error between the
sample average and the true average· and 3) the confidence
level that the allowable error is not exceeded. The number of
replicate tests i , howe er, often ba cd on customary practice.
For example, in acceptance te ring, ACI 3 1 8 considers a test
result as the average compressive trength of two molded
cylinder . Therefore, in correlation te ting, two replicate tan­
dard compre sion te t can be a umed to be adequate for
measuring the average compre!> ive strength at each level.
The nwnber of companion in-place tests at each strength
level should be chosen so that the averages of the in-place
te ts and compre sive trengths have imilar uncertainty. To
achieve this condition, the ratio ofthe number ofte ts should
equal the square of the ratio of the corresponding ingle­
opcrator coefficient of variation. If the number of replicate
compression test at each strength level is t� o the required
number of replicate in-place te ts is
(5.2.3)
where n ; is number of replicate in-place tests; V, i coefficient
of variation of in-place test; and Vs is coefficient of variation
of standard test.
For planning purposes, the coefficients of variation given
in hapter 4 can be u ed for the in-place tests. For molded
cylinder prepared cured, and te ted according to A TM
standards the single-operator coefficient of ariation can
be a sumed to be 3 percent (A TM C39/39M ). For core ,
a value of 5 percent may be assumed (A TM C42 C42 M ).
5.2.4 Regres ion analysis-After the data are obtained, the
trength relation hip should be determined. The u ual prac­
tice i to treat the average alues of the replicate compre -
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
sive strength and in-place test results at each strength level
as one data pair. The data pairs are plotted using the in-place
test value a the independent value (or X variable) and the
compre sive strength a the dependent value (or Y variable).
Regre ion analy is i performed on the data pair to obtain
the be t-fit estimate of the strength relation hip.
I L istorically, mo t strength relation hip have been
a swned to be traight line , and ordinary lea t- quares
(OL ) analy i ba been u ed to e timate the corre ponding
lope and intercept . The u e of OL i acceptable if an
estimate of the uncertainty of the strength relationship is not
required to analyze in-place test results such as if the proce­
dure in 7.2. 1 and 7.2.2 are u ed. If more rigorous methods,
such a those in 7.2.3 and 7.2 .4 are used to analyze in-place
te t re ult a procedure that i more rigorou than OL
should be u ed to e tablish the strength relationship and its
associated uncertainty.
The limitations of OLS analy is arise from two of its
underlying a umption :
(a) There i no error in the X value and
(b) The error ( tandard de iation) in the Y alue i con tant
xcept for mea ured maturity indexes the fi rst of the e
asswnptions is violated because in-place test (X value
generally have greater single-operator variability than
compression test ( Y value). In addition, it is generally
accepted that the ingle-operator variability of standard
cylinder compre ion test i described by a con tant coef­
ficient of ariation (ACI 2 1 4R) . Therefore the standard
deviation increa es with increasing compre sive trength,
and the econd of the aforementioned assumptions is al o
iolated.
a re ult, OL
aualy i will undere timate the
uncertainty of the trength relation hip (Carino 1 993). There
are, however, approaches for dealing \ ith these problems.
First, the problem of increasing tandard deviation with
increa ing average trength i discussed. I f test results from
group that have the same coefficient of variation are trans­
formed by taking their natural logarithms the tandard devia­
tion of the logarithm values in each group will have the same
value (Ku 1 969). For example, if the coefficient of variation of
a group of numbers, expressed a a decimal, equal 0.05, the
tandard de iation of the tran formed values will be approxi­
mately 0.05. Thus, th econd a un1ption of OLS can be
ati fied by p rforrning regre sion analy i using the average
of the natural logarithms of the test re ults at each trength
level. If a linear relationship is used, it fom1 is as fol lows
In
= a -r B lnl
( 5 .2.4a)
where In i the average of natural logarithms of compre sive strengths; a i intercept of line; B is slope of l ine; and
lnl is average of natural logarithms of in-place test re ults.
By obtaining the antilogarithm of ln , Eq. (5.2.4a) can be
tran formed into a power function
( 5 .2.4b)
The exponent B determines the degree of nonlinearity
of the power function. If B
I , the trength relation hip
=
is a straight line pa sing through the origin with a slope
25
=
A . If B =F I , the relationship has positive or negative curva­
ture, depending on whether B is greater than or less than I .
Regres ion analy is u ing the natural logarithms of the test
result pro ides two benefit :
( I ) atisfics an underlying as umption of OL analysis
(constant error in Y value), and
(2) llo� s for a nonlinear strength relation hip if such a
relationship is needed
Use of the transformed data implies that concrete strength
i distributed a a lognormal raU1er than a normal distri­
bution. It has been argued that, for the usual variabil ity of
concrete strength, the possible errors from this assumption
are not significant ( tone and Reeve 1 986).
ext, a method for dealing with the problem of error in the
X alues is discussed. Fortunately, regression analysis that
account for X error can be performed with linlc additional
computational effort compared with OL analysis. One uch
procedure wa proposed by Mandel ( 1 984) and wa u ed
by tone and Reeve ( 1 9 6) to de elop a rigorou procedure
to analyze in-place te t re ult (7.2.3). Mandel' approach
invol es the use of a parameter A defined as the ariance
(square of the standard de iation) of the Y ariable di ided
by the variance of the X variable. For the correlation-te ting
program the alue of A i obtained from the standard devia­
tions of the a erage compressive strength and in-place test
result . If the numbers of replicate for compre sive te ts
and in-place test arc cho en so that average alue arc
mea urcd with comparable preci ion, the value of A should
be c lo c to I .
The parameter A and the correlation te ting re ult -that
is, the a erages of the logarithms of the in-place results (X
alue ) and the average of the logariU1ms of compre si e
strengths ( Y alues}-are used to determine the strength
relationship using the calculation outlined in A.2 (Appendix
). The calculations involve the usual sum of squares
and cross-product used in OL analysi (Mandel 1 984).
The procedure is well suited for application on a personal
computer ith a pread hcet program.
Figure 5.2.4 i a graphical rcpre entation of the differ­
ence between O L analysi and Mande l ' pro edure. In
OLS analy i , the be t-fit traight line i the one that mini­
mizes th sum of quare of ilie vertical deviation of ilie
data points from the line, as hown in Fig. 5.2.4(a). Mandel's
analysis minimize the sum of squares of the de iations
along a direction inclined to the straight line, as shown in
Fig. 5.2.4(b). The direction of minimization depend on the
alue of A, which in turn depends on the ratio of the errors
in the Y and X values. As the error in the X value increa e ,
the value of A decrea e and the angle () in Fig. 5.2.4(b)
incrca e . An important feature of the Mandel analy is is
that the e timated tandard de iation of the predicted alue
of Y for a ne� alue of X account for error in the new X
alue and the error in the trength relationship {A.3 ).
I n ummary, regre ion analysis hould be performed
u ing the natural logarithm of the te t result to establi h
the strength relationship. This approach will accommodate
the increase in single-operator variabil ity with increasing
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
26
(a)
Q.)
15
ro
·.::::
ro
>
>-
X - Variable
(b)
Q.)
.0
ro
·.::::
ro
>
>-
X - Variable
Fig. .- A-Direction ofe!Tor minimi:ation in: (a) ordin(lly
lea t- quare (OLS) analy i : and (b) Mandel ' procedure
(Carino 1993).
trcngth. U ing a traight line to reprc ent the relation­
hip between logarithm alue i equivalent to as uming
a power function trength relation hip. The power function
can accommodate a nonlinear relation hip, if nece ary. To
be rigorou , the regre ion analysi procedure hould account
for the uncertainty in the in-place te I re ult (X error). Failure
to account for the X error will undere timate the uncertainty
of future e timate of in-place compre ive tTength. Thi
rigorou pro edure, howe er, i ju tified only when an equally
rigorous method will b u cd to interpret in-place te t re uh
( hapter 7 ); otherwi e, OLS analy i i a ceptable.
5.2.5 Procedures for correlation te ling I deally, it i
de irable to determine the compre i e trength and the
in-place te t re uh on the ame pecimen o that companion
te t re uhs are obtained at the ame maturity. Unfortu­
nately, thi i only po sible with those method that are truly
nondestructive, such a pulse velocity and rebound number.
For method that cau e local damage to concrete, eparate
pecimen are needed for obtaining compre sive trength
and the in-place te t re ult. Tn uch ca e , it i important
that companion pccimen arc tc ted at the arne maturity.
Thi i e pccially critical for early-age te t when trength
at a given age depend highly on the thermal hi tory. The
problem ari e becau e of difference in early-age tempera­
ture in pecimens of different geometric . An approach for
-
moderating temperature differences is to cure all specimens
under laboratory conditions in the same water bath.
Altemati ely, intemal temperatures can be monitored and
test ages adjusted so that compression tests and in-place
test arc performed at equal val ues of the maturity index.
Fail ure to perform companion tc ts on specimen that arc
at equal maturity will rc ult in an inaccurate strength rela­
tion hip that ill cause y tematic errors (or bia ) \ hen it
i used to e timate the in-place trength in a structure. The
following recommendations should be used in correlation
testing programs.
5.2.5. 1 Rebound number-At least 1 2 standard cylinders
should be cast. At each te t age, a set of I 0 rebound numbers
TM 805/C 05 M ) should be obtained from each pair
(
of cylinders held fitmly in a compres ion testing machine or
other suitable de ice at a pressure of approximately 3 M Pa
(500 p i). Rebound tc t for the de elopmcnt of a strength
relationship should be pcrfonned in the same orientation a
the tc t to be made on the stntcture. However, for hammers
ba ed on the ratio of hammer peed before and after impact,
the ori ntation of the instrument i not a factor. The cylinders
should then be tested in compression. I f it is not feasible to test
the cylinders with the hammer in the same orientation that will
be used to test the structure, the correction factor supplied
by the equipment manufacturer can be used to account for
differences in otientation. A mentioned in 3 .2, the urface
produced by the material of the cylinder molds can differ from
the urfacc produced by the fom1 material for the structure.
This factor should at o be considered in the correlation testing.
I f con id rable difference i exp cted between the surface of
the tructure and the cylinders additional prismatic pecimen
should be prepared for rebound tests. These specimens should
be formed with the amc type of forming material that wiU be
used in con truction, and they should be similar in izc to the
cylinders so that they will experience similar thermal histo­
ries.
hen the rebound number is dctctmincd on these speci­
mens, en urc that the specimen arc rigidly supported o that
they do not move during tc ting. l f thc pccimcn move, lower
rebound nwnbcr will be recorded and the trcngth relation­
ship � ill be bia cd.
For accurate c timatcs of in-place trcngth, the moisture
content and texture of the urfa e of the cylinder at the
time of the correlation te t hould be similar to tho e antici­
pated for the concrete in the structure at the time of in-place
testing. Practically, the only easily reproducible moisture
condition for concrete surfaces is the saturated condition.
5.2.5.2 Penetration resistance-For the probe penetra­
tion te t, at least 1 2 standard cylinders and a test slab large
enough for at lea t I probe penetration te ts should be cast.
For in-place testing of vertical element , the recommended
procedure i to cast a wal l specimen and take cores next to
the probe te t . All te t pecimen hould b cured under iden­
tical conctitions of moi ture and temperature. t each te t age,
two compre ion tests and three probe penetration tests should
be made. The recommended minimwn thicknes for the te t
lab is 1 50 mm (6 in.). The minimwn pacing between probe
penetrations is 1 75 mm (7 in.) and the minimum distance
from a probe to a slab edge i 1 00 mm (4 in.).
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
For the pin penetration test, it may be possible to perform
penetration tests on the side of cyl inders and ub equently
test the cylinders for compressive strength. arino and
Tank ( 1 9 9) howed that the urface damage produced by
pin penetrations into I 00 x 200 mm (4 x
in.) cylinders
did not re ult in strength reductions. omparative test ,
however, were not perfonned on pecimen with concrete
strength les than 25.5 M Pa (3 700 psi). Until further tudies
are conducted to confinn that pin penetrations do not affect
the compre ive trength of cylinder for a wide range of
concrete strength it i recommended that lab specimens be
used for pin penetration tests. A minimum of six penetra­
tion reading hould be performed at each test age. Di card a
result when it is obvious that an aggregate particle or a large
air void wa penetrated. In addition, according to
TM
03/ 803 M, if the range of penetration values exceeds 1 .6
mm (0.064 in.), the result with the maximum deviation from
the average hould be d iscarded and a new test performed.
Individual penetration hould b spaced between 50 and
1 50 mm (2 and 6 in.), and the minimum di tance from an
edge should be 50 mm (2 in.).
5.2.5.3 Pullout test- everal techniques ha e been u ed.
Pullout inserts have been cast in the bottom of standard
cylinder , and a pullout test \ as pe1fonned before testing
the standard cyl inder in compression ( Bickley 1 982b). I n
this c a e the pullout test i
topped when the rn a ·imum
load (indicated by a drop in the load with further di place­
ment) ha been attained. The insert i not extracted and the
cylinder can be capped and tested in compression. Alterna­
t ively, companion cylinders ha e been ca t with and without
in ert and the pullout te t has been performed on one
tandard cylinder and the other cylinder te ted in compre sion. I nvestigator have had problems \ ith both procedures,
particularly at high strengths, because radial cracking occurs
at the end of the cyl inder containing the pullout in ert. This
cracking is believed to re ult in lower ultimate pullout loads.
A third alternative ha been to ca t tandard specimen for
compression testing and to place pullout insert in cubes (or
slab or beams) so that the pullout test can be made in the
companion pecimen when the standard specimens are tested
( out os ct al. 2005). The latter approach i the preferred
method, providing con olidation i con i tent between
the tandard pecimen and the cube or other specimens
containing the pullout in erts, and the maturity of all speci­
mens at the time of te ting is the same. The recommended
minimum size for cubes i 200 mm (8 in.) when 25 mm ( I
in.) diameter inserts are used. Four inserts can be placed in
each cube one in the middle of each vertical side. For each
te t age, two standard cylinder hould be te ted and eight
pullout te ts perfonned. The same procedure applies to po !­
installed pullout test . Install the in erts on the same day that
pullout te t will be done.
5.2.5.4 ltrasonic pul e veto ity- l t i preferable to
develop the trength relationship from concrete in the struc­
ture. Tests should be performed on cores obtained from the
concrete being e aluated. Tests with tandard cylinder can
lead to unreliable relationships becau e of different moi ture condition between the cylinders and the in-place
27
concrete. The relation hip data should be obtained from a
testing configuration that is similar to the one used in the
field becau e the geometry of the test specimen may affect
the determination of the pulse velocity. The recommended
procedure is to select certain areas in the tructure that repreent d ifferent levels of pulse velocity. At the e locations it is
recommended that five elocity determination be made to
obtain a repre entative average alue of the pul e elocity.
For each mea urement, the tran ducer houJd be un oupled
from the urfa e and then recoupled to avoid y tematic eiTor
due to poor coupling {A TM C597). Then obtain at least
two cores from each of the same locations for compressive
trength testing. Pulse velocity measurement on these cores,
once they have been removed from the structure, will usually
not be the same a the velocitie mea ured in the structure and
are not representative of the pulse velocity of the structure.
5.2.5.5 Maturity method-The following procedure is
given m
TM I 074.
Prepare cylindrical concrete specimen according to
TM C 1 92/C 1 92 M u ing the mixture proportion for
the concrete intend d for the tructure. Embed temperature
sensors at the centers of at least two specimens. onnect the
ensors to maturity instrument or to a uitable temperature
recording device{s).
Moi t cure the specimen in a water bath or in a moist
room meeting the requirements of A TM 5 1 1 . Perform
compression test in accordance with A TM 39 39M at I ,
3, 7, 1 4, and 2 day . Test at lea t two pecimens at each age.
At each test age, record the average maturity index for
the in trumented pecimen . Plot the average compre si e
tr ngth as a function of the average maturity index. Draw a
be t-fit curve through the data. A uitable empirical equation
may be fitted to the data u ing least-square curve fitting.
Refer to Carino (2004) for possible equations. The resulting
curve is the trength-maturity relationship to be used for
estimating in-place trength.
5.2.5.6 ast-in-p/ace cylinder-If nece sary, test re ults
hould be corrected for the height-diameter ratio using the
alue given in
TM 42/C42 M . o other correlation is
needed because the specimens represent the concrete in the
placement and the test is a uniaxial compression te t.
5.3--E xisting construction
5.3. 1 General-There is often a need to evaluate the
in-place strength of concrete in existing structures. For
example planned renovation or change in the use of a struc­
ture may require detennination of the concrete strength for
an accurate a sessment of structural capacity. There also
may be a need to evaluate concrete strength after a struc­
tural failure, fire damage, or environmental degradation
has occurred. ometimes, errors or unforeseen conditions
occur during new construction and an e aluation i needed
to re ol e question about concrete trength. These itua­
t ions are similar because the need to detennine the in-place
strength of the concrete was not preplarmed. In-place testing
methods can be helpful in these evaluations.
In-place tests can be u ed in two ways to evaluate
existing construction. First, they can be used qualitatively
American Concrete Institute - Copyrighted © Material - www.concrete.org
28
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
to locate those portions of the structure where the concrete
appear to be different from other portion . In this case the
in-place tests can be u ed without a trength relationship
for the concrete in the tn1cture. The main purpo e of the
in-place testing is to establish where core should be taken
for strength determinations and other pertinent tests ( Cl
43 7R ). The rebound number and the pulse velocity method
are widely u ed for thi purpo e. e ond in-place methods
can be u ed for a quantitative a e rnent of the trength. I n
this c a e , a trength relationship m u t be e tabli hed for the
concrete in the tructure. The relationship can be developed
only by pe1forming in-place tests at selected locations and
taking companion cores for strength testing. Thus, the use of
in-place testing does not eliminate the need for coring but it
can reduce the amount of coring required to gain an under­
standing of the variations of trength in a structure, and it can
give a higher degree of confidence that the core taken truly
represent the condition being investigated.
5.3.2 Developing
trength relation hip-Typically
in-place te ting for evaluation of exi ting construction i
not preplanned, o the techniques that have traditionally
been used are ultra onic pulse velocity, rebound number, and
probe penetration. I n the nited Kingdom, the pull-off test is
al o used ( Long and Murray 1 984; Murray and Long 1 9 7).
The pul l-off test involves bonding a teel disc to the concrete
surface and measuring the resultant force required to pull off
the di c . In candinavia and other part of Europe, a post­
in tailed pullout test is widely used ( Peter en 1 9 4, 1 997) .
In current orth merican practice, the pull-off method is
not u ed routinely to e timate in-place compressive trength;
thu , this application i not con idered in the remainder of
this report.
For some te t methods, cet1ain factors hould be consid­
ered when testing exi ling structures. For example, for
surface tests (rebound number, penetration resistance, and
pull-off), pay special attention to those factors that could
affect the near-surface strength, such as carbonation moi ture content or surface degradation from chemical or phy ical proce e . ttrface grinding may be nece ary to expo e
concrete that represent the concrete within the tructure.
To de elop the tr ngth relationship, it i generally
nece my to correlate the in-place te t param ter with the
compre sive strength of cores obtained from the tructure.
In selecting the core locations, it is desirable to include the
w idest range of concrete strengths in the structure that is
possible. Often, rebound numbers or pul e velocity values
are determined at points spread over a grid pattern estab­
li hed on the area being e aluated. When the data are plotted
on a map contour l ine can be ketched in to outline the
variations in the concrete quality ( Murphy 1 984). Ba ed
on this initial survey six to nine different locations hould
be elected for coring and mea urement of the in-place te t
parameter. At each location a minimum o f t\: o core hould
be obtained to establish the in-place compre sive trength.
The number of replicate in-place tests at each location
depends on the test method and economic considerations, a
di cussed in Chapter 6. Because at least 1 2 cores are recom­
mended to develop an adequate strength relation hip, the use
of in-place testing may only be economical if a large olume
of concrete is to be evaluated.
ores hould be tested in a moi ture condition that is
repre entative of the in-place concrete. The recommended
procedure i to wipe off excess drill ing water, allow the
core to urface dry and place the cores in sealed plastic
bags. Refer to
TM 42/ 42M for additional guidance on
the handling and testing of cores.
After the average and tandard deviation of the in-place
te t parameter and core strength are detennined at each te t
location, the strength relationship is developed using the
same approach as for new construction ( 5.2.4).
In evaluating the a erage and tandard deviation of the
replicate in-place re ults, the recorded values hould be
checked for outlier {A TM 1 7 ). In general te t re ults
that are more than two standard deviations from the average
hould be scrutinized. Outliers may occur due to an improp­
erly performed test or a local ized. abnormal condition. I f an
obviou cau e of the outlier i identified that result hould be
ignored and the average and tandard deviation re alculated.
CHAPTER 6-I MPLEMENTATION OF I N-PLACE
TESTING
6.1 -New construction
6. 1 . 1
Preconstmction
con. ensus Before starting
con truction of the components of the tructure that are
to be tested in-place, a meeting hould be held among the
parties who are involved. The participants typically include
the owner, on tmction manager, structural engineer, testing
company general contractor ubcontractor (such a form­
work ontractor or po Hen ioning contractor) and concrete
supplier. The objective of the preconstruction meeting i to
clarify the test procedures to be used, the access require­
ments, the criteria for interpretation of test data, and the inter­
action among the partie . mutual understanding among the
involved parties will reduce the potential for dispute during
construction.
The meeting hould achie e a con en u on the following
critical is ue :
(a) Agreement on type of fonnwork material that will be
u ed becau e it may affect the correlation te ting
(b) The test procedure{ ) to be u ed, number and loca­
tions of tests, the acces requirements for testing, and the
assistance to be provided by the contractors in preparing and
protect i ng te t location and te ting equipment
(c) The criteria for acceptable test results for perfonning
critical operation , uch as fonn removal, post-tensioning
removal of re bores, or termination of accelerated or initial
cunng
(d) Procedures for providing acce s and any modifications
to formwork required to faci litate te ting
(e) Procedure and re ponsibilitie for placement of
te ting hardware, where required, and protection o f te t ites
( f) Procedures for the timing and execution of te ling
(g) Reporting procedures to provide timely information to
site personnel
American Concrete Institute - Copyrighted © Material - www.concrete.org
-
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
29
method as discussed in 5 .2 .3 . onsider, however, the prac­
ticality of the number of replicate tests; othenvise, in-place
testing programs will be avoided because of the financial
burden. Table 6. 1 .3 li ts the minimum number of individual
determinations per test location. A lo> er number is recom­
mended for those in-place te t methods that require installa­
tion of hardware compared with tho e methods that do not.
6. 1 .4 Providing acce s to test location -To perform
in-place te t during con truction, it i nece ary to provide
acce to the hardening concrete. The pecific detail will
depend on the test method, the type of tructural component,
and the type of formwork. Test locations should be selected
to avoid reinforcing steel. Finally, note that water absorp­
tion characteristics of the fonn surface at the location of the
in-place testing might affect the re ults of surface tests, uch
as the rebound number and pin-penetration method . Fonn
material for the in-place te t specimen in the correlation
testing must be similar to those u ed in construction.
For te ts on the soffit of labs formed with plywood, an
acce configuration a hown in Fig. 6. 1 .4a can be u ed.
circular hole i cut in the form and the plug that i cut
is attached to a backup plate that is temporarily fastened to
the fonnwork with crews. Test hardware, such as a pullout
in ert. is attached to the removable a sembly. When a test
i to be performed test hardware, if it exists, is loosened
and the backup plate and plug are removed to expose the
te t urface. To provide a mooth te t utface, a sheet metal
plate can be attached to the plug. A sealant should be used
to seal the gap betv een the plug and backup plate to prevent
leakage of fresh cement pa te. The diameter of the plug will
depend on the pecific pacing requirement for the te t
method a di cu ed in 6. 1 .6, and it should provide at lea t
25 mm ( I in.) of clear space around the perimeter of the
plug to avoid testing concrete near the edge of the plug. For
access through metal forms, a similar backup plate assembly
can be fabricated of metal plate. A typical access configura­
tion for u e on the ertical surface of a metal form is shown
in Fig. 6. 1 .4b.
The acce s types hown in Fig. 6. 1 .4a and 6. 1 .4b are appli­
cable to all the in-place testing methods except for the matu­
rity method and ca t-in-place cylinders. Figure 6. 1 .4c i llus­
trate typi al technique for installing maturity meter . The
dispo able mini-maturity meters can be in erted directly into
(h)
pproval procedures to allow construction opera­
tions to proceed if adequate strength is shown to have been
achieved
( i ) Procedures to be followed if adequate strength i not
shown to have been achieved
6. 1 .2 umber of test locations The te t should provide
a reliable mea ure of the trengtb of the te ted component at
the time the test are made. Therefore, sufficient test locations
should be provided so that there are ufficient te I re ult
to adequately characterize the concrete trength within the
pmtion of the tructme being e aluated. The tenn "test loca­
tion"' means a region on the structure where an in-place test
procedure i to be executed. At a te t location, one or more
single or replicate in-place test may be performed.
The number of test location
hould account for the
fol lowing considerations:
(a) Becau e te ts will be performed at early age � hen
strength gain of concrete depend highly on temperature the
initial test may how that adequate strength ha not yet been
achieved. It will then be nece ary to top testing after the
initial te t have been made and to rete t at a later age. uffi­
cient test locations have to be provided to allow for repeat
tests and to satisfy the criterion for number of te t required
to allow critical operation to proceed.
(b) I f tests are made at ages under 1 2 hours after the
concrete is cast it i expected that the in-place strength will
ha e high variability due to variations in temperature at the
test location . In thi case, increa e the number of provided
test location by I 0 to 25 percent.
Tables 6. 1 .2a to 6. 1 .2d provide recommendations for
testing variou tructural component . For each te t method,
the table how:
(a) The number of te t locations or acces points that
should be provided per stated volume of concrete
(b) The minimum number of test locations that should
be available for stati tical analysis to determine concrete
strength.
The number in these tables are based on experience
con idering the criticality of the tructural component and
practical ity.
6. 1 .3 Number of te. ts per location-The number of
in-place te t to be petformed at a te t location could, in
theory be determined ba ed on the repeatability of the te t
Table 6.1.2a-Recommendations for suspended slabs, shear walls, and core walls*
•
umber of test locations provided
Te>t method
First 75 m 1
Rebound number
( I 00 yd1)
umber of locations to test
Each additional 1 5 m1 (20 yd1)
First 75 m1 ( 1 00 yd1)
Each additional 1 5 m1 (20 yd1)
2
10
I
I
6
I
20
Probe penetration
Pin penetration
15
2
10
I
Pullout
15
2
10
I
Ultrasonic pul e velocity
15
2
10
I
5
2
5
I
-
I
5
I
-
Mantrity
--
Ca t-in-place cylinder'
•corr walls that typ•cully surround elc.,ator halls arc u ually located at the
' l'or >lab on ly.
enter of a bu•lding
and fom1 the >tructural backbone of the build10g.
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
30
Table 6.1.2b-Recommendation for other walls per 1 50 m2 (200 yd2)
--
N umber of test location provided
Wall thinner than 300 rum
Test method
( I ft)
Rcbow1d number
20 to 25
Probe penetration
Ptn penetration
1-
-+---
to 1 0
-+---
Pullout
Ultrasonic pulse veloeity
- --
1 0 to 1 5
15 to 20
10
to 1 2
5
Probe penetration
5 to 8
5
Pin penetration
5 to 8
5
Pullout
5 to
6
'ltra onic pu1 e \ eloeity
5 to
6
5
5
•Recommendations are based on
the assumpt ion s that there are 6 to
each test area and that ea h olumn contams approximately I
Grenter numbt
test nrca
o f t c t; 'hould
bt provided and te
contains more t han I 0 columns.
8 to 1 2
10
1--
columns m
m1 ( t .5 yd1) of concrete.
Minimum number of
locations provided
loeations to test
Rebound number
6 to 9
5
Probe penetration
6 to 9
5
Pin penetration
6 to 9
5
Pullout
6 to 9
6
(.,1 trasonic pulse veloeity
6 to 9
6
Maturity
5
5
-
-
•Recommendations apply to the number of t�t locauon> provided/tested before
rcmo> I of form
and
agam before application of construction loading from ne.•t
5
Tc t method
Minimum number of locations to test
Rebound number
10
Probe penetration
Pin pcoctrati o
Pullout
ltrasonic pul
Test method
8
5
r-
Table 6.1.2d-Recommendations for columns with
spandrel beams per 40 m3 (50 yd3)
I--
10
8
---
Table 6.1.3-Number of replicate tests at each
location
ted for larger columns or "'here the
umber of test
10
--
Minimum number of
8
Maturity
10
to 1 2
Test method
5
thicker
6
5
Rebound number
Wall 300 mm ( I fl) tbi k or
6 to
Table 6.1.2c-Recommendations for individual
columns*
LO
Wall thinner than 300 mm
( I ft)
5
umber of test
( I fl) thick or
umber of loeation to test
thicker
-- --
! O to 1 5
Maturity
mm
- --
1 0 LO 1 5
--
WaiL 300
-
level
of con truction It is assumed that corbel . if present. arc cast integrally with columns
or spandrel beams.
the top mfaccs of labs, or they can be embedded deeper
into the slab using a cup-lid assembly to avoid interference
with finishing operations . The cup may al o be placed within
openings on the ide of vertical form . For electronic matu­
rity meters temperature probe arc in crtcd into the structural
clements. For meter with reusable probe , the u ual practice
i to embed an expendable plastic tube into the frc h concrete
and to place the probe within the tube (Fig. 6. 1 .4c(b)). A
thermal couplant, which i a type of grease, hould be applied
to the probe before in crtion into the tube to ensure accurate
measurement of the concrete temperature. For meters that u c
thcimocouplc wire as en ors, the wire arc fastened to rein­
forcing bars before concreting. After testing is completed, the
r
vcloei ry
I
I
Maturity
Ca t-in-place cylinder
3
6
I
2
l
l
2
thennocouple wires are cut flu h with the concrete u1face,
and the exec s wire can be reu.ed.
ast-in-place cylinders do not requjrc pecial accc proviions. The upporting le ve for the cylinder mold is nailed
directly to the formwork. It i only nece ary to en ure that the
top urface of the specimen \ ill coincide \ ith the top urface
of the slab. I f the top of the pecimen i too lo , it will be diffi­
cult to locate and extract the cylinder. If the top of the spec­
imen i too high, fini hing operations will di rupt the molds.
6. 1 .5 Distrihution of te. Is-Te t location should be
di tributed throughout the component being tested o that
the results provide an accurate indication of the trcngth
di tribution within the component. In electing the te ting
location con idcration should be given to the most critical
location in the structure in term of trcngth requirement
( uch a po !-ten ioning tre ing location ) and expo ure
condition ( uch as lab edge ), e pecially during cold
weather. When a large number of tests are required for stmc­
tural components such as slabs distribute the test locations
in a regular pattem. For test method that require few tests
such as ca t-in-place cylinders, choose location that are
critical in each concrete placement.
For te t on vertical members such a column , walls and
deep beam the vertical location within the placement is
important. For vertical members, there is a tendency for the
concrete trength to be higher at the bottom of the placement
than at the top of the placement. The magnitude ofthi varia­
tion is influenced by many factor , such a mixture compoition, type and degree of consolidation, aggregate shape,
and environmental condition ( Murphy 1 984; Munday and
Dhir 1 984; Bartlett and MacGregor 1 999). It is impossible
to predict accurately the magnitude of trength variation
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
(a)
M1n1mum diameter
Inserted 1nto
top surface
depends on
method
Hole (1f needed)
31
Installed In
formwork
using cup
Minimum limits to
avo1d obstruction of
Fig. 6. 1. 4a-Access for use on vertical swfaces and soffits
with woodenforms.
·.·...
( b)
----tl� To meter
Plastic
:
Column,
beam, or .,
wall
·
Butterfly bolt
Hole (if needed)
Steel back-up plate
Fig. 6. 1. 4c-lnstallation of maturity meters illlo fresh
concrete: {a) disposable mini-meter: and (b) sen or of elec­
n·onic meter.
6.2-Ex isting construction
6.2. I Prete ling meeting-
Fig. 6. 1. 4b-Acces for u e on vertical swface and soffits
with steelforms.
expected in a given component. AI o, code-writing commit­
tee have not addre ed the e trength ariation . A a re ult,
engineering judgment i needed in planning and interpreting
the re ult of in-place te t on vertical member , particularly
when testing member with depth greater than 300 mm
( 1 2 in.). imilar engineering judgment will al o need to be
made when te ting deep slab ection .
6. 1 .6 Critical dimensions-Te t uch a rebound number,
penetration re i tance, and pullout produce orne urface
damage to the concrete, and te t re ult are affected by the
condition within the zone of influence of the pecific te t.
A a re ult, the A TM tandard pre cribe minimum dimenion to a sure that test re ults are not influenced by neigh­
boring te t , specimen boundarie , or reinforcing teel . Te t
location hould be po itioned to conform with the dimen­
ional requirement in Table 6. 1 .6.
di cu ed in 5. , there
are many reason for detem1ining the in-place trength f
concrete in e i ling tructure . In-place testing i orten
one facet of an overall inve ligation to e tabli h tructural
adequacy. The guideline in AC T 437R should be followed
to de clop the complete plan of the inve ligation and iden­
t i fy other a peel of the field tudy to complement concrete
trength determination.
The plan for the in-place te ling program will depend
on the purpose of the inve tigation. A prete ting meeting
hould be held among the member of the team who share a
common intere t in the te t re ult . t the conclu ion of the
meeting, there hould be a clear under tanding of the objec­
tive of the in c tigation and agreement on the re pon ibili­
t ie of the team member in acquiring the te t data and the
procedure for obtaining and analyzing te t re ult . When
acce to the concrete for te ting i re tricted by architectural
co ering , detailed plan hould be developed to accompli b
thi acce .
6.2.2 Sampling p/an-111 de eloping the te ting program,
con ideration should be given to the mo t appropriate
ampling plan for the pecific ituation. A TM
823/
823M provide guideline for developing the ampling
plan . Although the tandard deal primarily with the dri lling
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
32
Table 6.1.6-Dimensional requ i rements for
in-place tests according to ASTM standards*
Tes1 mclhod
Requiremenl
Minimum dimen ion
Rebound number
Th1 knes of member:
1 00 mm (4 in.)
Diameter of test area:
00 rnm ( 1 2 in.)
Minimum distance
Between lesl pomt :
25 mm ( I in.)
Minimum di. ranee
Between probes: 1 75 1111n (7 in.)
To edge ofconcre1e: 1 00 mm (4 in. )
Minimum disrnnce
Pin penetration
Between pin : 50 nun (2 in.)
To edge of c n rete: 50 mm (2 in.)
Marimum disrnn e
Between pin : I 0 mm (6 in.)
Miuimum clear spacing
Pullout
Be1ween in en : 1 0 time in en head diame1er
To edge of member: Four limes head diamcler
From edge of failure urfacc 10 reinforcing bar: ne
in en head diameler or ma imum aggregate ize.
' hichcver i larger
•The cu.,nl ' cr>�on of lhc A TM lc 1 methods hould be consullcd before plannong
m-ploce 1es1s 10 ensu"' 1ha1 proper >pacing and clearance requ�rcmenJS ""' sausfied.
of cores or awn amples, there i a section addres ing
in-place testing.
In general two sampling ituations can be encountered.
In one ituarion all the concrete is believed to be of imilar
compo ition and quality. For thj case, random ampling
hould be pread out o er the entire tructure and the result
treated together. onsult A TM E l 05 to w1derstand the prin­
ciples of random sampling. The structure should be parti­
tioned into different regions and a random number table used
to determine objectively which area to te t. Objecri e random
ampling is neces ary to apply probability theory and make
valid inferences about the propertie of the population (all the
concrete in the tructure) ba ed on the ample te t re uJt .
The second sampling ituation ari es
hen available
infonnation uggest that the con rele in different ection
of the tmcture may be of different compo it ion or quality,
or when the purpo e of the in estigation i to examine fai lure
or damage in a specific section of a tructure. In this case,
random sampling should be conducted within each ection
of the structure where the concrete is uspected of being
nominally identical . Te t results from different ection of
the structure should not be combined unless it i shown that
there are no tatistically significant differences between the
a erage test re ult in the different section .
6.2.3 umber of test As discu ed in 5.3 , the in-place
te ting program for an existing tructure involve two
pha e . Fir t, the trength relation hip must be e tabli hed
by testing drilled core and mea uring the corresponding
in-place test parameter near the core location . The loca­
tions for correlation testing hould be cho en to provide a
wide range in concrete strength. s mentioned in 5. 3 .2, a
minimum of six to nine te t location hould be selected for
obtaining the correlation data. In general cores hould be
drilled after the in-place tests are performed. At each loca­
tion, two cores should be dril led, and the following number
of replicate in-place tests should be performed to provide
the average value of the companion in-place test parameter:
Tes1 mclhod
Rebound nwnber
I
Replicates at each location
10
Probe or pin penelrnlion
3 to 6
Ultrasomc pulse veloc1ty
5
Pullout
3
The number of replicate in-place tc t i ba cd on con id­
cration of the single-operator ariability of the method
and the co t of additional te ting. For example the ingle­
operator repeatabil ity of the ultra onic puJ c velocity te t
i low, and the cost of replicate reading at one location i
low. Therefore, five replicate reading arc recommended to
ensure that a rcprcsentati c value will be obtained bccau c of
the variability in the efficiency of the coupling of the tran duccr to the tructurc. Ln makjng the replicate pulse velocity
dctcnnination . the transducer should be moved to nearby
location to evaluate the area where cores will be taken. The
dimcn ional requirement pre cntcd in Table 6. 1 .6 hould be
observed for all test methods.
The c ond phase of the in-place tc ting program involve
performing the in-plac tests at other locations and c 6mating the compre sive strength ba cd on the trcngth
relationship. The number of test loca6on for thi pha c
will depend on several factors. First, there arc the statis­
tical factors. According to the principle set forth in A TM
1 22, the number of tc t depend on the variability of the
concrete trcngth the acceptable error between the true and
ample average, and the acceptable ri k that the error will
be exceeded. mong the e factors, the variabil ity of the
concrete is a predominant factor in determining the number
of required tc t . For a given ac cptablc error and I vel of
ri k, the nwnber of te t increases with the quare of the
ariability (
TM 1 22).
Economic considerations also influence the tc ting plan.
For orne ca c , the co t of an cxtcn ivc investigation might
outweigh the economic benefit. Because the co t of an inve tigation i related to the amount ofte ting performed, a high
degree of confidence, due to a large ample ize is obtained
at a higher cost. The election of a te ling plan involve trad­
cotr between economic and degree of confidence.
CHAPTER 7-I NTERPRETING AND REPORTING
RESULTS
7.1 -General
Standard tatistical procedures hould be used to interpret
in-place test . I t i not ufficient to simply average the values
of the in-place tc t results and then compute the equivalent
compressive strength by mean of the pre iou ly e tabli hed
trength relation hip. It is nccc ary to account for the uncer­
taintie that exist. While no procedure has been tandard-
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
ized for determining the tenth-percentile in-place strength
based on the results of in-place te ts, proponents of in-place
testing have developed and are using statistically based
interpretations.
Four stati tical methods for evaluating in-place test re ults
are revie\ ed in the following section . The fir t two methods
are irnilar and are ba ed on the idea of tati tical tolerance
factors. The e two method are simple to u e, requiring
only tabulated tati tical factor and a calculator. Becau e
of their underlying a umptions, howe er, the tati tical
rigor of these method bas been questioned. As a result,
more rigorous methods have been proposed. The rigorous
method are more complex and require an electronic preadheet or computer program for practical implementation.
Table 7.2.1 a-One-sided tolerance factor for 1 0
percent defective level (Natrella 1 963)
umber of
tests 11
Column I
1--1--
1-�
-
7.2.2 General tolerance factor method (l lindo and
Bergstrom 1 9 5)-The acceptance criteria for trength of
concrete cylinder in ACl 2 1 4 R are ba ed on the a ump­
tion that the probability of obtaining a test with strength less
than .fc' i le s than approximately I 0 percent.
suggested
method for evaluating in-place te ts of concrete is to deter­
mine the lower tenth percentile of trength, with a pre cribed
confidence level.
I t has been e tabli hed that the variation of cylinder
compressive strength can be modeled by the normal or the
lognormal distribution function, depending on the degree
of quality control. In ca e of excellent quality control,
the di tribution of compre i e trength re ult i better
modeled by the normal di tribution; in ca es of poor control,
it is better modeled by a lognonnal distribution ( H indo and
Bergstrom 1 985).
In the general tolerance factor method, the lower tenth­
percentile compre sive strength Y0.10, is estimated from
4
5
-1-1-
--
6
7
Confidence level
75%
olumn 3
2. 0 1
4.2
2. 1 34
1 . 96 1
1 .860
--
-
1 .7 9 1
-
f-
9
1 .702
10
1 .67 1
II
1 .646
12
1 .624
13
14
15
20
-·
---f--
95%
90%
olumn 2
1 .740
7.2-Statistical methods
7.2.1 Danish method (Bickley 1 982h)
Thi method ha
been de eloped for analysis of pullout test results. The
pullout strengths obtained from the field test are convert d
to equivalent compre sive trength by mean of the trength
relationship (correlation equation) determined by regre sion analysis of pre iously generated data for the particular
concrete being used. The standard deviation of the converted
data i then calculated. The tenth-percentile compre sive strength of the concrete is obtained by subtracting the
product of the standard deviation and a statistical factor K
(which varie with the number o f te ts made and the de ired
level of confidence, p) from the mean of the converted data.
Although Bickley ( 1 982b) did not state it explicitly, the
stati tical factor is a one-sided tolerance factor (Natrclla
1 963 ), a dis u ed further in 7.2.2. The K factor for
different number of te t and a p
75 percent are given in
olwnn 2 ofTable 7.2. l a. The example in Table 7.2 . 1 b illu trates how the Danish method is applied. The first column
shows the equivalent compressive strengths corresponding
to the I 0 individual pullout te t re ults. The econd column
hows the values and calculation used to obtain the tenth
percentile strength at p - 75 percent. The example use 1 0
te t rc ults, but another appropriate number may be used in
larger placements.
33
f
tI
I
I
I
· �t
3. 1 7
2.742
Column 4
-
-I-
-f-
--
6. 1 5
-
4. 1 63
3.407
2.494
3.006
2.333
•.
2.2 1 9
2.582
2 . 1 33
2.454
2.065
2.355
2.0 1 2
2.275
1 .966
2.2 1 0
155
-
1 .92
-
--
2. 1 55
-
1 .5 9 1
I. 95
1 .577
1 .866
1 .528
1 .765
1 .926
1 .702
1 .838
1 .657
1 .778
1 .623
1 .732
25
1 .496
30
1 .475
35
1 .458
40
1 .445
50
1 .426
I
+
I
I
-
-t-
2. 1 08
2.068
1 . 59
1 .697
1 . 560
1 .646
-l
-
in-place test results by con idering quality control, number
of tc ts n, and the de ir d confidence lc cl p for the c ti­
matcd trength. Three quality control level arc con idcrcd:
excellent, average, and poor, with the di tribution function
of trength as umed a normal, mixed normal-lognormal,
and lognormal, respecti ely. uggested alue of p are 75
percent for ordinary structures, 90 percent for very important
buildings, and 95 percent for crucial parts of nuclear power
plants ( H indo and Bergstrom 1 985). Thus by selecting a
different p-value the user can adjust the le el of conserva­
tism in estimating in-place trength that is consi tent with
the criticality of the project.
The tolerance factor K, the sample a crage Y, and tandard
de iation r arc u cd to e tablish a lower tolerance limit­
that is, the lower tenth-percentile trcngtb . For a normal
distribution function, the estimate of the tenth-percentile
strength Yo.t o can be determined a follows
Yo. to = Y - Ksy
( 7.2.2a)
where Yo. 1 o is lower tenth-percentile of strength ( I 0 percent
defective); Y i
ample average trcngth; K is one- idcd
tolerance factor (Table 7.2. l a); and sy i
ample tandard
de iation.
The tolerance factor i determined from tati tical cbarac­
teri tics of the normal probabil ity di tribution and depend
on th number of test n, the confidcn e level p, and the
American Concrete Institute - Copyrighted © Materi al - www.concrete.org
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
34
Table 7.2.1 b-Example of Danish method
Individual equivalent compressive
slTenglh. MPa (psi)•
0.9
alculalions
27.5 (3990)
0.8
25.0 (3620)
24. (3550)
25.0 (3620)
22.5 (3260)
24.0 (34 0)
25.5 (3700)
28.5 (4 1 30)
Mean Y = 25.7 MPa (3730 psi)
vt o
landard dc ali n s1 = 2.3 MPa
(330 psi)
K - 1 .67 1 t
Tcnlb-pcr clllilc
lrcng l h Y Ks t·
= 2 1 .9 M P a ( 3 1
0 psi)
25.0 (3620)
30.0 (4350)
"Convened from pullout force measurements using trength relarionship
'The •alues of the constant K for the 75 percent confidence level are given in Column
2 of Table 6. 1 .2a.
defect percentage. Value of K are found in reference book
such a that by atrella ( 1 963 ). Table 7.2. 1 a pro ide one­
sided tolerance factors for confidence levels of 75, 90, and
95 percent and a defect le el of I 0 percent.
For the lognormal distribution, Y0. 1 0 can be calculated in
the arne manner; use the average and tandard deviation of
the logarithm of trength in q. (7.2 .2a).
By dividing both ide of Eq. (7.2.2a) by the average
trcngth Y, the fol lowing i obtained
(7.2.2b)
where Vy is coefficient of ariation (cxpres cd a a decimal).
In Eq. (7.2.2b), the tenth-percentile trcngth i exprc ed
a a fraction of the a erage trength. F igure 7.2.2 i a plot of
Eq. (7 .2.2b) for p 75 percent and for coefficient of varia­
tion of 5, I 0 1 5, and 20 percent. Thi figure how that a
the ariability of the te t re ult increa es or as fewer test
are performed, the tenth-percentile strength is a maller frac­
tion of the a erage strength.
The tolerance factor method i
imilar to the Dani h
method. The re ults of the in-place tests are converted to
equivalent compressive strengths using the strength relation­
ship and the equivalent compres ive trcngth are used to
compute the ample average and tandard de iation.
The example in Table 7.2.2 illustrate the application of
the tolerance factor method for probe-penetration te t . The
que tion in the example i whether the in-place strength of
concrete in a lab is ufficient for the application of post­
tensioning, if the compressive trength requirement for post­
ten ioning i 20 M Pa (2900 p i). The numbers in the fir t
column are the mea ured expo ed length of each of eight
probes, and the econd column gi es the corresponding
compressive strengths ba ed on the previou Iy e tabli hed
strength relationship for the concrete being evaluated. For
eight te t andp 75 percent, the tolerance factor i 1 .74. It i
a umed that the nonnal di tribution de cribe the variation
=
=
:
t
t
+
5
10
15
·- - · · · · · ·
:· · - - · · · · · · ·
... ....... --------·
y
__.£..!£.
y
0.7
· · · ·--····
� cv�ii ;ci
; cv=o os
1
··········-
·
l
�
20
25
-
·
· - - - - - - - - - - - - - - - - - --- cv=o 1s
: cv=o 20
0.6
0.5
0.4
0
Number of Tests
30
Fig. 7. 2. --Ratio of tenth-percentile trength to average
strength a a fim lion ofcoefficient ofvariation and number
ofte f (normal distribution as wned).
of concrete strength. Thu , by ub tituting the cocffi icnt of
variation and the tolerance factor into q. (7.2.2b), the ratio
of Y0. 1 0 t o the average trength i 0 . 3 . Therefore, Y0 10 i
I .6 M Pa (2700 p i). Becau e the tenth-percentile trength
i Je than 20 M Pa (2900 p i ), po t-ten ioning hould not be
applied. Thu , additional curing time i needed. Refer to 4. I
for a di cu ion of data interpretation for tructure under
con !ruction.
7 .2.3 Rigorous method (Stone and Ree1·e 19 6)- The
pre eding method con crt each in-place te t re ult to an
equivalent compres ive strength alue by mean of trength
relation hip. The average and tandard deviation of the
equivalent compre i e trength are u ed to compute the
tenth-percentile in-place slTength. Two major objection
have been raised to the e method ( tone el al. 1 9 6· tone
and Ree e 1 986):
(a) trenglh relation hip i presumed to have no error
(b) Variability o f the compre i e trength in the tructure
i a umed to be equal to the ariability of the in-place tc 1
rc ult
The fir t factor will make the e timate o f in-pia e tenth­
percen tile trength uncon er alive wherea the econd
fa tor will make the e timale over- on ervative.
tone and Reeve ( 1 9 6) de eloped a comprehen i e
technique for tali tical analy i of in-place te I re ult that
attempted to address the perceived deficiencie of the toler­
ance factor method . Only a general ummary o f the method
i given herein. Thi rigorou method encompa e the
following procedure :
( I ) Regrc ion analy i
to c labli h the
lrcngth
relation hip
(2) E timating the variabil ity o f the in-place compre i e
trength ba ed on the re ult of the correlation le 1 and te t
on the tructure
( 3 ) alculating the probability di tribution of the e ti­
mated in-place, tenth-percentile strength
(4) For the rea ons gi en in 5 .2.4, logarithm of the te I
result are used in the analy i , and lrength relation hip
i a umed to be a power function. Regre ion analy is i
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
Table 7.2.2-Example of general tolerance factor
method
trength relationship:
Y ( MPa) = - I + 0.69L (mm)
( Y (psi) = - 1 45 + 2540L [in. ) )
Exposed length L. mm ( in.)
Compressive strength Y. M Pa ( psi)
30 ( 1 . 1 )
1 9.7 (2850)
3 - ( 1 .3 )
23.2 (3360)
34 ( 1 .34)
22.5 (3260)
-
-
23.2 (3360)
35 ( 1 .3 )
-
38 ( 1 .50)
36 ( 1 .42)
3 1 ( 1 .22)
30 ( 1 . 1 8)
*
25.2 (3660)
23.9 (3460)
20.3 (2950)
1 9.7 (2850)
Mean ( Y) = 22.2 M Pa (3220 psi ).
tandard deviation (st ) - 2. 1 MPa (300 psi).
Coefficient of variation CVt) = 9.3 percent.
For 11 = 8 and 75% confidence level: K = I. 74.
Y0 10 = ( 1 - K V1 ) Y = ( I - 1 .74 x 0.093) x 22.2 = 1 8.6 MPa (2700 psi).
performed using the Mandel procedure discussed in 5 .2 .4
and A.2. The errors associated with the best-fit strength rela­
tionship arc used to estimate the in-place tenth-percentile
strength at any desired confidence level.
A novelty of the rigorou method is the approach u cd to
c timatc the variability of the in-place compressive trcngth.
In haptcr 4 it i shown that the single-operator ariabi lity
of in-place test result is generally greater than compressive­
test rc ult , which i why objection ha c been raised against
a surning that the ariability of the in-place compr ssi c
strength equals the variability of the in-place tc t result .
ln the rigorous mctl10d, it i as umed that the variability
of compressive strength divided by the variability of the
in-place test results i a constant. Thus, the ratio obtained
during correlation testing is assumed to be valid for the test
conducted in the field. This provides a mean for e timating
the variability of the in-place comprcs ivc trengtb based on
the results of the in-place te ts (7.2.4).
The in-place tenth-percentile strength computed by the
rigorou procedure account for the error as ociatcd with
the tr ngth relationship. The user can determine th tenth­
percentile strength at any desired confidence level for a partic­
ular group offield test results. I n addition, the user can choose
the percentile to be a value other than the tenth percentile.
tone ct a!. ( 1 9 6) computed the tenth-percentile strengths
by the rigorous method and compared them with those
computed by the Oani h and tolerance factor method .
The c calculations u cd imulatcd in-place test data having
different mean value and standard de iations. lt was found
that for an assumed confidence lc el, the strengths estimated
by the Danish and tolerance factor methods were lower than
the vaJucs ba cd on the rigorous method. The difference
were a high as 40 percent when the in-place tests had high
variability (coefficient of variation = 20 percent). Compared
with the rigorou method, the Danish and tolerance factor
35
methods give more conservative estimates of in-place
compressive strength, but they do not appear to provide a
consistent confidence level. One reason for the inconsis­
tency of the tolerance factor method is the assumption that
the variability of the in-place compressive trength i the
same as the ariabi lity of the in-place test results. Experi­
mental field tudie arc needed to compare the in-place,
tenth-percentile trength e timated by the e method with
the value obtained from many core te t . Only then can the
reliability of these methods be e aluated.
7.2.4 Alternative method (Carino 1 993)-The rigorous
method developed by tone and Reeve ( 1 986) has not
received widespread acceptance among concrete technolo­
gists because of its complexity. arino ( 1 993) proposed
an altemative method that retain the main feature of the
rigorous method but can be implemented easily with spreadheel oftware.
TI1e basic approach of the alternative method i illustrated
in Fig. 7.2.4. The Mandel procedure (a outl ined in .2 )
i u ed to obtain the trength relation hip from correlation
data. The result of the in-place test and the strength rela­
t ionship are u ed to compute the lower confidence limit of
the estimated average in-place sn·engtb at a desired confi­
dence level. Finally the tenth-percentile strength is deter­
mined assuming a lognonnal distribution for the in-place
concrete strength. alculations are performed u ing natural­
logarithm value .
In the following paragraphs the procedure for estimating
the in-place strength is explained further. When the in-place
trength is to be estimated, repl icate test are performed on
the structure. The average of the logarithm of the in-place
te t i u ed to compute the logarithm of the average in-place
compressive strength using the strength relationship
Y = a + bX
(7.2.4a)
where Y is logarithm of the e timated average in-place
compres ive strength; X is average of the logarithm of the
in-place test performed on the structure· and a,b are inter­
cept and slope of the trength relationship.
Next, the lower confidence l imit for the e timated average
trength i computed. Thi lower l imit i obtained u ing Eq.
A.3) for the tandard deviation sr of an stimated alue of
Y for a new X. The lower confidence limit for the average
concrete strength is as follows
( 7.2.4b)
where Y1,"' i lower confidence limit at confidence level
a; 1111. La is tudent's /-value for m- 1 degree of freedom
and confidence level a; and 111 is the number of replicate
i n-place te ts.
Table 7 .2.4 li t Student t- alue for m - 1 degrees of
freedom and ri k (or confidence) levels of 5 and 1 0 percent.
The choice of risk level depends on the criticality of in-place
concrete strength in the overall assessment. When strength is
critical, a lower risk level such as 5 percent, should be used.
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
36
4./
.....
Q.)
"0
c
>.
()
-
0
E
......
.I::.
·;::
ca
C)
0
....J
/
/
/
/
/
/
��
Table 7.2.4-Student's t-values for m-1 degrees of
freedom and risk levels of 0.05 and 0.10 (Natrella
1 963)
/
Too�
m-1
e/
,(._ Lower
Confidence
Limit (Y�ow )
Average
2
2.920
3
2.353
-
4
2. 1 32
-
6
-
-
Logarithm of I n -Place Test Result
Fig. 7.2. 4-A/ternative method to estimate compressive
strength ba ed on in-place te t (Carino 1 993).
The di tribution of in-place compres i e strength i
described by a lognonnal di tribution, and the tenth-percen­
tile trength i computed as fol lows
5
7
-----
To
L886
- --
2.0 1 5
1 .943
1 .895
10
1 .638
1 .533
1 .476
- --
1 .440
1 .4 1 5
8
1 .860
1 .397
9
1 .833
1 .383
10
I. 1 2
1 .372
II
1 .796
1 .363
12
1 .782
1 .356
13
L771
1 .350
-
14
-
16
-
15
-
-----
1 .76 1
1 .753
- --
1 . 345
I. 4 1
1 .746
1 .337
17
1 .740
1 .333
18
1 .734
1 .330
19
L729
1 . 328
-
-
-
-
-
(7.2.4c )
where Y0_10 i logarithm of trength expected t o b e exceeded
by 90 percent of the population; and s,r i tandard deviation
of the logarithms of concrete strength in the structure.
The value of Scf i obtained from the a sumption ( tone
and Reeve 1 986) that the ratio of the tandard deviation of
compressive strength to the standard deviation of in-place
te t re ult has the same value in the field a " as obtained
during the laboratory correlation te ling. Thus, the following
relation hip is assumed
(7.2.4d)
where cfi c1 are standard deviation oflogarithm of compre sive strength in the structure and laboratory re pectively;
and sx, il are tandard deviation of logarithm of the in-place
results in the structure and laboratory, re pectively.
The final step i to convert the re ult obtained from Eq.
( 7.2.4c) into real unit by taking the antilogarithm.
A c lose examination of the alternative procedure hows that
the average compressive strength estimated by the strength
relation hip ( q. (7.2.4a)) i reduced by two factors. The
fir t factor, which is given by Eq. (7.2.4b), accounts for the
uncertainty of the strength relationship and the uncertainty
of the average of the in-place test result . The second factor
which i given by Eq. (7.2.4c) accounts for the ariability
of the in-place compres ive strength. Thus it is believed that
the alternative procedure strikes a balance between stati tical rigor and practicality of u e.
mentioned, the proce­
dure is well uited for implementation using a computerized
preadsheet or a pecialized computer program ( hang and
gives examples
mino 1 99 ). ection A of ppendix
that compare the e timated in-place trength using the toler­
ance factor and alternative method .
7.2.5
umma
With the exception of ca t-in-place
cylinder test , in-place test provide indirect mea urc of
concrete strength. To arrive at a reliable e timate of the
in-place strength, the uncertaintie involved in the e ti­
mat mu t be con idered. Thi ection ha di cu ed orne
techniques de eloped for thi purpo e. The tolerance factor
methods discus ed in 7.2. 1 and 7.2.2 ha e been used succes fu l ly in the analysis of pullout test data. Therefore, they may
be adequate for test methods that ha e good correlation with
compressive strength, uch as the pullout te t.
The tolerance factor methods however do not account
for the main source of uncertainty in a rational way. This
has led to the development of more rigorou procedures
a di cus cd in 7.2.3 and 7.2.4. The c new methods arc
designed to pro ide reliable c timatcs of in-place trcngth
for any tc 1 procedure. The c rigorou method , how vcr,
should be incorporated into ca y-to-usc computer program
for practical usc.
7.3-Reporting results
Report forn1 for the different tc ts and different purpo c
will vary. A variety of report form arc appropriate. U ually
relevant
TM tandard de cribc the information required
on a r port. Where in-place tc ring i made at early age ,
omc pccific reporting data arc de irablc.
ct of fonn ,
imilar to those developed by an engineer for u c in pullout
testing, i shown in Fig. 7.3a to 7.3c. These can serve a
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
8
A
c
D
E
F
G
H
K
ABC TESTING COMPANY LETTERHEAD
l'rojcct 1\o._____
@)
@) @
® ®
®
r-
J
@) @) @
:
@T 9 W @ ® ®
:�
Repon
licnt:
Pour #2
t T :J @
37
ddress
2
3
_
_
_
_
_
_
Tc�tmg of In· PI cc
Pr �e t
Anent ion:
Dear
_
o. _
_
(in
sequence)
tn:nglh
_
,me:_
_
_
_
ddrc
ir:
The follow ing
the above site.
an: th
results
Locauoo in structure:
Pour.
Proposed time
fonn removal
Test
of in-place
tests
of
MPa concrete at
Individual
tests n:o.ults
_
_
_
_
_
_
_
Dme
Time
(MPo)
of
R�ul!li Summa ry
Number of tests made.
Fig. 7.3a-Example of form used to identifY locations of
in-place tests in a floor slab ofmultistOiy building.
Mean in-pi:JCC-strcnl!.th (MPa )
Standard
dcviation.!MI'a):
Minomum in-place stren!!,th (MPa):
ABC TESTING COMPANY
Field Record of I n-Place
Test
Number
Test
result
Estimated
compressove
1
strength
2
3
4
Testing
Project Number
Remar�
Requorements of
__ MPa mean and __
before stripping and reshoring nrelare n
Your5 vel) trul).
b
IPa mmimum strength
the abo\e rc ul�.
Copy gn·en
Dtttc:
Time:
Project Name
Location in
structure
to srte upermtendcnt
--:�-..,.--,---­
igned -
(}11' f4'VI
J
« C'f
AJ
(/r1 Co
n
l
m
d
o
r
;-
Placement
Date
5
1 met
Time
6
7
Size
Mix No
Fig. 7. 3
Curing Conditions
made at very early ages and the time to complete a place­
ment is long, there could be a significant age-strength aria­
tion from start to finish of the placement.
(b) Record offield-test result. (Fig. 7. 3b)-Thi i the
form on which test data, the calculated re ults, and other
pertinent data are recorded at the site. The form shm n i n
Fig. 7 . 3 b has been designed for evaluating the data with
the Danish or tolerance-factor methods ( minimum strength
is the tenth-percentile strength). It includes pro isions for
entering information on maturity data, protection details and
concrete appearance to corroborate the te t data during cold
weather. Due to the critical nature of formwork remo al, a
recommended procedure is for the field technician to phone
the data to a control office and obtain confirmation of the
calculation before gi ing th re ult to the contractor.
(c) Report of test result. (Fig. 7. 3c)- This fom1 i u ed
to report the in-place test results. The example shown in
Fig. 7.3c is a multicolor, self-carbon form designed to be
completed at the site by the technician, with copies given
to the contractor's and structural engineer's representatives
when the results have been checked. lt provides for identifi­
cation of the placement involved, the individual results, and
the calculated mean and minimum strengths. It record the
engineer's requirements for form removal and states if the c
requirements have been met. It requires the contractor's
representative signature on the testing company's copy.
ampleform for reporting in-place test results.
8
9
10
11
12
13
14
15
Maturity __•c-h
Temperature at tesL
Ambient __·c
Within enclosure __· c
Appearance of top surface:
Ca l cu l ations
:
Number of tests: ___
Esttmated strength:
Mean:
Standard devtation:
K-Value:
Mimmum Strength:
Mean - (K • sd):
Remarks:
Technooan
Cl"oecl<ed by
lnsuument number
Fig. 7.3b-Sampleformfor on- ite recording a_(in-place te I
results.
useful models for developing fom1s to report the re ults of
other in-place tests.
Briefly, the three form provide for the fol lowing:
(a) Record oftest locations (Fig. 7.3a)- Thi form gives a
plan vic\ of a typical floor in a specific multi tory building.
TI1e location of each test is noted. The location of maturity
meter , if installed, can also be hown. Location data arc
important in case of low or ariablc results. Where test arc
American Concrete Institute - Copyrighted © Material - www.concrete.org
38
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
CHAPTER 8-IN-PLACE TESTS FOR ACCEPTANCE
OF CONCRETE IN N EW CONSTRUCTION
8.1 -General
Traditionally, acceptance te ting for new construction has
been limited to judging acceptabi lity of the concrete deliv­
ered to the project on the ba i of slump air content, and
compressive strength. cceptable con rete that i placed
con ol idated, and cured ac ording to standard of good
practice will perfonn according to de ign as umption .
Exceptions occur when there is clear evidence of inadequate
consolidation or d istress, such as cold joints and excessive
cracking, or when inadequate protection was provided in
cold weather.
The durabil ity of exposed structure depend trongly on
the curing hi tory of the concrete. Therefore, it is desirable to
have a surance that the concrete in the fini hed tructure ha
the neces ary propertie to attain the de ired level of perfor­
mance. i n-place t sting offer the opportunity to obtain thi
a urance \ hen u ed a a component in a comprehen ive
quality a urance program. The Great Belt L i nk project in
Denmark is one of the first large-scale construction proj­
ects in which the owners relied on in-place testing (pullout
tests) instead of tandard laboratory trength tests to assess
the acceptabil ity of the concrete layer protecting the rein­
forcement (Vincent en and l lenrik. en 1 992 ). This major
construction effort erve a a model for future project
where in-place quality assurance is important.
In orth America, there i a reluctance to abandon tradi­
tional acceptance procedures that ha e erved their purpose.
In-place te ting however, offers th opportunity to le n the
reliance on te ting of tandard-cured cylinder a the ole
method to judge acceptability of concrete delivered to the site.
The added benefit of in-place testing is that it provides assur­
ance that the finished construction has the propertie speci­
fied by the designer. This chapter di cu se the potential for
in-place te ting a an alternative tool for acceptance te ting.
8.2-Acceptance criteria
The following reviews the current acceptance criteria in
ACI 3 1 8 and propo es how in-place testing may be u ed
a an altemati e to te ting standard-cured cylinder in new
con truction.
8.2 . 1 Molded cylinders-According to
C l 3 1 M- 1 4
I 3 1 - 1 4), the evaluation and acceptance of concrete
are based on tests of cylinder molded at the job site and
subjected to standard laboratory curing in accordance with
A TM
3 1 /C3 1 M . ection 26. 1 2.3(b) of
I 3 1 M- 1 4
(A I 3 1 - 1 4) tates a follows:
"Strength level of a concrete mixture hall be acceptable if
( I ) and (2) are satisfied:
( 1 ) Every arithmetic average of any three con ecuti e
trength te t equal or exceed.fc'
(2) o trength te t fall below .fc' by more than 3.5 M Pa
(500 psi) i f.fc' is 35 M Pa 5000 psi ) or les ; or by more than
0. 1 O.fc' i f.fc' exceed 35 M Pa ( 5000 p i ).'
In addition, according to 26.5.3 .2(d) of
I 3 1 M- 1 4
(A 1 3 1 8- 1 4), the building official or l icen ed de ign profe -
(aCii
sional m ight require le ting of field-cured cylinders to check
the adequacy of curing and protection of the concrete in
the tructure. The acceptabi l ity of curing and protection, a
indicated by the field-cured cyl inder strength , is defined in
ection 26.5 . . 2(e)
"Procedures for protecting and curing concrete shall be
con idcrcd adequate i f ( l ) and (2) arc sati ficd:
(I)
erage strength o f field-cured cylinder at test age
designated for determination of.fc' i equal to or at lea t 5
percent of that of companion tandard-cured cylinder
(2) verage strength of field-cured cylinder at test age
exceeds .fc' by more than 3.5 M Pa (500 psi)."
8.2.2 ores- I n the event that a strength te t of tandard­
cured cylinders i more than 3.5 M Pa (500 p i) below /. ',
A I 3 1 - 1 4 requires step. be taken to ensure adequacy of
the tructure. ore may have to be drilled to verify the
in-place trength. Three core are required for each strength
test fail ing to meet the specified criteria. In j udging the
acceptability of the core trength , ection 26. 1 2.4. 1 (d) of
ACI 3 1 8- 1 4 tate the fol lowing:
" oncrete in an area represented by core te t shall be
considered structurally adequate if ( I ) and (2) are satisfied:
( I ) The average of three core i equal to at lea t 5
percent of.fc'
(2) o single core i les than 75 percent off..'."
8.2.3 In-place tests-Based on the aforementioned require­
ment for judging the acceptabil ity of in-place concrete in
ne\ con truction based on core strengths, the following
a ceptance criteria based on in-place testing are propo ed:
The concrete in a tmcture is acceptable if: I ) the esti­
mated average, in-pia e, compre i e trength ba ed on an
ASTM tandard in-place te t procedure equal at lea t 5
percent ofj..'; and 2) no te t result estimates the compre i e
strength to be less than 75 percent of.fc'.
In this ca e, an estimate of in-place strength at a test loca­
tion is the value of Y1.,w computed by Eq. (7.2.4b), and the
e timated average in-place strength i the average value of
Yt011 • Before these criteria can be put into effect, however,
standard practices need to be adopted for electing the
number of in-place tests to be done at one location and for
stati tical analy i of in-place te t data.
8.3-Early-age testing
The primary rea on for using in-place te ts in new
con truction is to determine \ hether it i afe to perform
critical operations, such as form removal or post-tensioning.
The in-place tests provide e timates of compressi e strength
at ages that are usually much earlier than the age for attaining
the pecified strength. The criterion frequently u ed to judge
the acceptabil ity of early-age strengths to permit critical
con truction operation
uch as formwork remo al, is that
the e timated in-place compre ive trength hould be at
lea t 75 percent of.fc'. In thi ca e th e timated lower tenthpercentile in-place strength, Y0.10• hould be compared to the
required strength. When such a requirement is specified,
early-age te ting may faci l itate final acceptance of concrete.
I n high-rise con !ruction, economic factors result i n acceicrated schedules in which critical operation may be planned
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
39
Table 8.3-Results of standard-cured cyl i nder and i n-place tests at 28 days (fc' = 30 M Pa [4350 psi])
Project
Project 2
I
Pullout tests
Pullout tests
4
Mean trength. MPa (psi)
tandard deviation
. M Pa {p t)
Range. MPa (psi)
Expected percentage of results
below/,.'
I
15
34.4t {4990)
{ 630)
35.91 {52 1 0)
3 .2 {5540)
2.7t (390)
3.9 {570)
2.71 (390)
3.5 ( 5 1 0)
30.5 to 44.5 (4420 to 6450)
29.9 to 40.5 (4340 to 6920)
32.5 to 40.51 (471 0 to 5 70)
30.9 to 43.5 {44 0 to 63 1 0)
l .63s
2.23s
4.9
ctual percentage of results
below/,.'
.
2 . 34s
1 .2
1 .4
1 .2
one
one
2. 1
"A resuh i the a•ernge of two-cylinder test or the average of two or more pullout tests
'Mean and standard deviat ion of estimated compres ive trength based on strength relationship.
as early as I to
day after concrete placement. To meet
the early-age strength requirement , the contractor may
choose to u e a concrete mixture that ill exceed the speci­
fied design strength. Experience has shown that requiring a
minimum trength of 75 percent of .fc' at early age ( I to 3
day ) will u ually en ure that the in-place trength will be at
lea t/,.' at 2 day , i f proper curing is u ed and the pecifica­
tion do not al low mixtures that achieve all their strength
gain at the time of form removal.
For example, for a specified design strength of 2 M Pa
(4000 psi), the in-place trength to permit form removal
may have to be at least 2 1 M Pa (3000 p i). A II owing for the
inherent ariation of concrete strength the a erage in-place
strength may ha e to be 25.5 M Pa (3700 p i) to ensure that
the early-age trength criterion is ati fied. In this example,
the a erage early-age, concrete trength ha to equal 93
percent of the pecified trength. Therefore, it i rea on­
able to a ume that if the early-age ( 1 - to 3-day trength
requirement is satisfied, then at 2 day the pecified de ign
strength will undoubtedly be achieved. For additional assur­
ance, in-place tests can be made on the structure at 2 day .
Bickley ( 1 9 4) reported on two demonstration projects
where in-place testing was u ed for early-age strength deter­
mination of horizontal elements, as wel l as for confirmation
of the 2 -day design strength. Permi sion to waive stan­
dard cylinder testing wa obtained from the building offi­
cial. I nnovati e project pecification defined the frequency
of in-place te t and the procedure to follow in doing the
test and reporting the re ults. Acceptance of the concrete
' as based on the results of pullout tests performed on the
structure at 2 days. For comparison, tandard-cured cylin­
ders were also tested at 2 day , but these strengths were
not reported. Table .3 ummarizes the result . The specified
design strength for both project wa 30 M Pa (4350 psi ).
Individual pullout te t result were converted to compres i e
strengths ba ed on the trength relation hips, and the e esti­
mated trength were used to compute th tat1 t1c hown
in the e ond and fourth columns of the table. Ba ed on the
standard deviation , the expected percentage of trength
below.f/ were computed. In all ca es, these percentage were
les than I 0 percent, which i the approximate alue impl ied
in A I 1 8. For both project , the in-place te t results clearly
howed that the concrete had acceptable trength.
Tn conclusion, current legal contract for the ale and
purcha e of ready mixed concrete are usually ba ed on the
28-day trength of tandard-cur d cylinder . For the time
being, therefore, the e cylinde have to be ca t . When
in-place test are made at an early age, however, the accept­
abi lity of the concrete can be as e ed at that time. If the
concrete is ati factory, there i no need to te t the tandard
cylinder . I f the early in-place te ts indicate a problem with
concrete in a particular placement, the related standard
cylinder are available for testing.
A
CHAPTER 9-REFERENCES
T committee document and document publi hed by
ther organization are li ted fir t by d cument number full
title, and year of publication followed by authored docu­
ment li ted alphabetically.
American
oncrete Institute (A I)
A I 2 1 4R- l l -Guide to Evaluation of trength Te t
Results of oncrete
A T 2 1 4.4R - l 0-Guide for Obtaining ore and Inter­
preting ompres ivc trength Re ults
A T 228.2R- I -Report on Nondestructi c Tc t Method
for E aluation of oncrerc in rructure
A I 30 1 - 1 6- pecification for Structural on rete
A I 306R- 1 6-Guide to old Weather Concreting
A I 308R- 1 6- uide to Extemal uring o f oncrete
A I 30 . I - l l - pecification for uring oncrete
A I 3 1 8- 1 4-Building ode Requirement for tructural
oncrete and ommentary
A T 3 1 8M - 1 4-Building ode Requirements for true­
rural oncrete and ommentary ( Metric)
A T 325 .9R- 1 5-Guidc for
onstruction of oncrete
Pa cments
A T 325. 1 1 R-O l - ccelerated Techniques for oncrcte
Pa ing
A T 437R-03- trength Evaluation of Exi ting oncrete
Building
American Concrete Institute - Copyrighted © Material - www.concrete.org
40
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
ACI 562- 1
Code Requirements for Assessment,
Repair, and Rehabilitation of Existing Concrete tructures
and Commentary
A
TM lmemational
European ommifleefor tandardi=ation (CE �
E
1 379 1 -2007-Asse sment of In- itu
ompre si e
trength in tructures and Precast Concrete omponents
lmemational oncrete Repair Institute {I R l)
A TM
3 1 /C3 I M- 1 8- tandard Practice for Making
and uring Concrete Test pecimens in the Field
A TM C39/39M- 1 8-Standard Te t Method for ompre sive trength of Cylindrical Concrete pe imen
A TM C42/ 42M- I
tandard Te t Method of Obtaining
and Te ting Drilled ores and awed Beams of oncrete
A TM
94/ 94M- 1 7- tandard Specification for
tandard Practice for Making
and uring oncrete Test pecimen in the Laboratory
A TM CS I I - 1 3- tandard
pecification for M ixing
Room, Moi t abinets, Moi t Rooms, and Water torage
Tanks Used in the Testing of Hydraulic Cements and
oncrete
A TM
597- 1
tandard Te t Method for Pul e
elocity Through Concrete
A TM
803/ 03M- 1 7- tandard Test Method for
Penetration Re istance of Hardened oncrete
A TM C 05/C805M - 1 3- tandard Test Method for
Rebound umber of Hardened oncrete
A TM
23/C823 M- 1 2(20 1 7)- tandard
Practice
for Examination and ampling of Hardened oncrele in
on truction
A TM
873/ 873M- 1 5
tandard Test Method for
Compres ive Strength of Concrete Cylinder
ast in Place
in Cylindrical Mold
A TM
900- 1 5- tandard Test Method for Pullout
trength of Hardened oncrete
A TM
9 1 8/ 9 1 8M- 1 3- tandard Test Method for
Mea uring arly- ge ompre sive trength and Projecting
Later-Age trength
I 074- 1 7- tandard Practice for E timating
A TM
oncrete trength by the Maturity Method
A TM C 1 5 3/ 1 583M- 1 3
tandard Tc t Method for
Tensile trength of oncrete urfaces and the Bond trengtb
or Ten i le Strength of Concrete Repair and Overlay Mate­
rials by Direct Tension ( Pull-off Method)
A TM E I 05- 1
tandard Practice for Probabi lity
amp l ing of Materials
A TM
1 22- 1 7- tandard Practice for Calculating
ample ize to Estimate, With pecified Precision, the
Average for a haracteristic of a Lot or Proce s
A TM E 1 7 - I
tandard Practice for Deal ing with
Outlying Ob ervations
Brili II Standards in titution (BSI)
B 1 88 1 -207: 1 992-Te ting concrete. Recommendation
for the
e ment of oncrete Strength by ear-to-Surface
Test
B
60 9:20 I
ssessment of I n-situ
ompressi e
omponents.
trength in tructure and Precast
omplementary Guidance to that Gi en in B E 1 3 79 1 .
I Rl 2 1 0.3 R-20 1 3 Guide to Using In- itu Tensile
PullolfTe ts to E aluate Bond of oncrete urface Materials
Authored documents
ACI ommittee 3 1 8, 1 9 3, "Building ode Requirement
for Reinforced oncrete (A I 3 1 -83) and ommentary
(3 1 R- 3 )," American Concrete Institute, Farmington H il ls,
M l , 266 pp.
Al-Manaseer
. and quino E. B., 1 999, " ind or
Probe Te t for onde tructive Evaluation of ormal and
High- trength Concrete," A I Materials Journal V. 96, o.
4, July-Aug., pp. 440-447.
Ballarini, R.; Shah, . P.; and Keer, L. M., 1 986, "Failure
haracteristic of Short Anchor Bolts Embedded in a Brittle
Material," Proceeding of the Royal Society of London.
Series A, Mathematical and Physical Sciences, . 404 I 26,
pp. 3 5 -54. doi : 1 0. 1 09 /rspa. l 9 6.00 1 7
Bartlett, . M . , and MacGregor, J. G., 1 999, "Variation of
In-Place oncrete trength in tructures," ACJ Materials
Journal, V. 96 No. 2 Mar.- pr. pp. 2 6 1 -269.
Berg trom, . G., 1 953, " uring Temperature, Age, and
trength of Concrete Magazine of oncrete Research, .
5
o. 1 4, Dec. pp. 6 1 -66. doi: 1 0. 1 680/macr. l 953 .5 . 1 4.6 1
Bickley, J . A., 1 9 2a, Concrete Optimization," Concrete
International, V. 4, o. 6, June, pp. 38-4 1 .
B ickley, J . A., 1 9 2b, " ariability of Pullout Te ts and
In-Place oncrete trength,' Concrete International, . 4,
o. 4, Apr., pp. 44-5 1 .
B ickley, J . A., 1 9 4, "The Evaluation and Acceptance of
oncrete Quality by In-Place Testing," in itu ondestructive Testing of Concrete, P- 2 . M. Malhotra, ed., Amer­
ican oncrete I nstitute, Fannington H i lls, M l pp. 95- 1 09.
Bloem D. L., 1 96 " oncrete trength in tructures,"
AC/ Journal Proceedings, Y. 65, o. 3 Mar., pp. 1 76- 1 7.
Bocca. P. 1 984, "Application of Pull-Out Te t to High
Strength Concrete trength E timation " Materials and
Structure , . 1 7, No. 3, pp. 2 1 1 -2 1 6.
Bogas, J. A.; Gome , M . G.; and Gome ,
., 20 1 3,
"Compre sive
trength E aluation of tructural L ight­
weight
oncrete by
on-destructive U l trasonic Pul e
elocity ethod." Ultrasonics, . 53, o. 5 , pp. 962-972.
doi: I 0. 1 0 1 6/j .u ltras.20 1 2. 1 2 .0 1 2
Breysse, D . 20 1 2, " ondestructive E aluation of
oncrete trength: An H i torical Review and A
ew
Per pecrive by ombining DT Methods," on /ruction
& Building Material , V. 33, pp. 1 39- 1 63 . doi : 1 0. 1 0 1 6/j .
conbuildmat.20 1 1 . 1 2 . 1 03
Brook . A. G.; chind ler,
. K.; and Bame , R.
2007, "Maturity Method Evaluated for Yariou Cementi­
tious Materials," Journal of Materials in
ivil Engi­
neering, V. 1 9, No. 1 2, Dec., pp. 1 0 1 7- 1 02 5 . doi : 1 0. 1 06 1 /
(
CE)O 99- 1 56 1 (2007) 1 9: 1 2( I 0 1 7)
'
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
Bungey, J . H., and Madandoust, R., 1 992, "Factors
Influencing Pull-off Tests on
oncrete," Magazine of
Concrete Research, V. 44, o. 1 5 , pp. 2 1 -30. doi: I 0. 1 680/
macr. 1 992.44. 1 5 .2 1
Bungey, J. H . · M i l lard . G.; and Grantham, M. 2006,
Testing of oncrete in tructure , fourth edition, Taylor &
Franci Group. London and ew York, pp. 339.
arette, G. G. and Malhotra, V. M . 19 4, "In itu
Te t : Variability and trength Prediction at Early ge
In Siw!. ondestructive Testing of oncrete P- 2, V. M .
Malhotra, ed., American oncrete I n titute, Farmington
H i l ls, M I , pp. 1 1 1 - 1 4 1 .
arino, . J. 1 984, "Maturity Method: Theory and Appli­
cation," ement, oncrere and Aggregares V. 6, No. 2, pp.
6 1 -73. doi: 1 0. 1 520/
A l 03 5 J
. J ., 1 993, " tatistical Methods to valuate
arino,
In-Place Te t Result ," ew Concrete Technology: Robert E.
Phil/eo Symposium, P- 1 4 1 , T. C. Liu and G. . I I off, eds.,
American Concrete i nstitute Farmington l lill , M l , pp. 39-64.
arino
. 1 . , 2004, "The Maturity Method," 1/andbook
on ondestructive Testing of oncrete, hapter 5, econd
edition, . M . Malhotra and . J. arino, eds., R Press
I nc. Boca Raton, FL.
Carino, . 1 . ; Lew, H . : and Volz, . K., 1 983a, "Early
Age Temperature ffect on Concrete trength Prediction by
the aturity Method," A I Journal Proceedings, V. 0, o.
2, Mar. -Apr., pp. 93- 1 0 I .
arino, . 1 . , and Tank, R. ., 1 9 9, " tati tical harac­
teristic of ew Pin Penetration Test Cement, Concrete
and Aggregates V. I I , o. 2. pp. 1 00- 1 0 . doi : 1 0. 1 520/
C I O I I OJ
arino
. J. and Tank R. ., 1 992, 'Maturity Function
for Concrete Made with arious ements and Admixture ,
A CI Materials Journal, V. 9, No. 2 , Mar.- pr., pp. I - 1 96.
arino, 1 . J . ; Woodward, K. A.; Leyendecker, E. V: and
Fattal . G. 1 9 3b, "Review of the kyline Plaza ol lap e,"
oncrete International, V. 5 No. 7, J uly, pp. 35-42.
habowski, A. J ., and Bryden- mith, D. W., 1 980
e ing the trength of oncrete of In itu Portland
cment Concrete by I nternal Fracture Te ts " Magazine
of Of/crete Research, . 32, No. 1 1 2, pp. 1 64- 1 72. doi:
1 0. 1 6 O/macr. I 9 0.32. 1 1 2 . 1 64
hang, L. M . , and Carino, . J., 1 99 , "Analyzing In-Place
oncrete Tests by omputer," oncrete Inrernational . 20,
o. 1 2, Dec. pp. 34-39.
Courard, L: B is annette, B.; arbacz, .; aysburd, A.;
von Fay, K.; Moczul ki, G . ; and Morency, M., 20 1 4, " ffect
of M isalignment on Pulloff Test Results: umerical and
Experimental sse ment ," ACI Materials Journal, V. I l l ,
o. 2. Mar.-Apr., pp. 1 53- 1 62. doi : I 0. 1 4359/5 1 6 645 1
Delatte
.; Will iamson M . ; and Fo\J ler D., 2000, "Bond
trength Development with Maturity of l l igh-Early trength
Bonded oncrete Overlay , A CI Materials Joumal. . 97,
o. 2, Mar.-Apr., pp. 2 0 1 -207.
Dilly, R. L., and Vogt, W. L., 1 9 , "Pullout Test Maturity, and P
pread heet oftware,'' ondestructive Testing,
P- 1 1 2, H . . Lew ed., American oncrete Institute, Farm­
ington H ills M I , pp. 1 93-2 1
."
"
"
"
41
Domone, P. L., and Castro, P. F., 1 987, "An Expanding
leeve Test for In- itu Concrete and Mortar trength valu­
ation " Proceedings tmctural Faults and Repairs 87 Engineering Technic Pres
dinburgh.
Facaoam 1 . , 1 970, " on-De tmctive Testing of oncrete
in Romania," Proceedings yrnposium on on- Destructive
Testing of oncrete and Timber, June 1 1 - 1 2, 1 969, I nstitu­
t ion of Civil Engineer London, pp. 39-49.
Facaoam, 1., 1 9 4, "Romanian Achievement in onde­
tmctive Strength Te ting of oncrete," In Situ ondestruc­
tive Tesring ofConcrere, P-82, . M . Malhotra, ed. Amer­
ican oncrete Institute, Fa1mington H ills M l , pp. 35-56.
Ferretti . 2004, "A Cell Method (CM) ode for odeling
the Pullout Test tep-Wise " Computer Modeling in Engi­
neering & ciences, V. 6, No. 5 Nov., pp. 453-476. ( ME )
Freiesleben-Hansen, P. and Pedersen J., 1 977 "Matu­
rity omputer for ontrolled uring and Hardening of
Concrete," ordi k Berong, o. I pp. 1 9-34.
Gallo, G., and Popovic , J. ., 2005 "The U e of Surface
Wave to E timate I n-Place trength of oncrete " Journal
ofAdvanced oncrete Technology, . 3 o. 3, pp. 3 55-362.
doi: 1 0.3 1 5 1 /jact.3.3 5 5
udra, T., and Stawiski B . , 2000, " on-destructive
trength
haracterization of Concrete Using
urface
o. I pp. 1 -6. doi :
Wave ," DT & E International, V. 33
I 0. 1 0 1 6/ 0963- 695(99)0002 -6
Hell ier A. K.; an alone, M: arino
. J . ; tone, W.
.; and l ngralfea, . R., 1 9 7, "Finite-Element Analysis
of the Pullout Test sing a onlinear Di crete Cracking
pproach,' Cement, oncrete and Aggregates, . 9. o. I ,
pp. 20-29. doi : 1 0. 1 520/ C 1 0393J
H i ndo K., 1 990 "In-Place Bond Te ting and urface
Preparation of oncrete," Concrete International, V. 1 2 o.
4, Apr., pp. 46-4 .
Hindo, K. R. and Bergstrom W. R., 1 985, " tati tical Eval­
uation of the In-Place ompressive trength of Concrete "
Concrete International, V. 7 o. 2, Feb., pp. 44-4 .
Hoerner T. E., and Darter M. 1 . , 1 999. " uide to Devel­
oping Performance-Related pecification for P
Pave­
ments Volwnc ! -Practical Guide. Final Report and
ppendix
( PRS for the Acceptance of Jointed Plain
oncrete Pavement)," Report
o. F//WA -RD-9 -155,
a hington D .
Federal H ighway Adrnini tration (FH A)
Ismail, M . P.; Yu of, K. M.; and Ibrahim, . . 1 996, "A
ombined Ultrasonic Method on the stimation of Compres­
sive Concrete
trength " lnsighr (American ociety of
Ophthalmic Registered Nurses), . 3 , No. I I , pp. 78 1 -7 5 .
Jones, R . , 1 949, "TI1e
on-destructive Testing o f
oncrete,' Magazine of oncrete Research, Y. I , o . 2 ,
June, pp. 67-7 . doi: 1 0. 1 6 O/macr. J 949. 1 .2.67
Jones, R. 1 962,
ondestructive Te ting of Concrete,
ambridge Univer ity Pre , London.
Keiller A. P., 1 9 2, "Preliminary In estigation of Te I
Methods for the A es ment of trength of In itu Concrete,
Technical Report No. 42.5 5 1 , Cement and oncrete Asso­
ciation, Wexham prings U K, 37 pp.
Khoo, L. M., 1 984, "Pul lout Technique- n Addi­
t ional Tool for I n itu oncrete trength Determination,"
American Concrete Institute - Copyrighted © Material - www.concrete.org
'
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
42
In itu/Nondestructive Testing of oncrete,
P- 2 , V. M .
Malhotra. ed., American oncrete In titute, Farmington
H i l ls, M I , pp. 1 43 - 1 59.
Kjellsen, K. 0., and Det\vi ler, R. J., 1 993, "Later- ge
trength Prediction by a Modified
aturity Model," A CI
Materials Journal . 90, o. 3. May-June pp. 220-227.
Krenchel, I 1 . , and Peter en . G., 19 4, "In-Place Testing
with Lok-Te t: Ten Years' Experience," Pre entation at
International onference on !11 itu/Nond stru rive Testing
of oncrete, Ottawa, Ontario, anada.
Ku, H. H. 1 969, " otes on the
e of Propagation of
Error Fommlas," Precision Measurement and alibration­
tati tical Concepts and Procedures, National Bureau of
tandards P 300 . I pp. 33 1 -34 1 .
Leshchinsky, A. M . , 1 99 1 , " ombined Method of Deter­
mining ontrol Measures of oncrete Quality." Materials
and Structures, . 24, o. 3, pp. 1 77- 1 4. doi : 1 0. 1 007/
BF02472983
Le hchin ky A. M: Yu, M . ; and Goncharova, . S.
1 990, " ithin-Te t Variability of orne Non-De tructive
Method for Concrete trength Determination," Magazine
of oncrete Research, V. 42, o. 1 53 pp. 245-24 . doi :
I 0. 1 6 O/macr. l 990.42. 1 53 .245
Le lie, J . R., and Cheesman W. J., 1 949, '·An ltrasonic
,
Method of Deterioration and racking in oncrete tructures
ACI Journal Proceedings, . 46 o. 9 ept. pp. 1 7-36.
Lew H. . 1 980, West Virginia ooling Tower ollap e
aused by Inadequate Concrete
trength,"' Ci il Engi­
neering, . 50, o. 2 pp. 62-67.
Long A. E., and Murray A. M . , 1 984, "The ' Pull-Off'
Pmtially De tructiv Te t for Concrete," in Situ ondestru five Testing of oncrete P- 2, . M . Malhotra, ed. Amer­
ican oncrete Institute Farmington H ill , M l , pp. 327-350.
Mailhot, G.; B isaillon, G . ; arette, G. .; and Malhotra,
V. M . , 1 979 "In-Place
oncrete trength:
e\ Pul lout
Methods A I Journal Proceedings . 76
o. 1 2 Dec.
pp. 1 267- 1 2 2.
Malhotra V. M. 1 97 1 , "Maturity oncept and the
timation of oncrete trength
Review," information
ircular o. I 277, Department of Energy, Mines and
Resource , Ottawa, 43 pp.
Malhotra, V.
. 1 975, ""Evaluation of th Pull-Out Te t
to Detennine trength of In- itu Concrete," Materials and
Structures . , o. 43, pp. 1 9-3 1 .
Malhotra, V. M., 1 976, "Testing H ardened
oncrete:
onde tructive Methods,' A I Monograph o. 9, American
oncrete In titutellowa tate niversity Pres Farmington
H i lls, M I , 204 pp.
Malhotra V. M. and arette, G. G. 1 980, " ompari on of
Pullout Strength of oncrete with Compressive trength of
Cylinders and ores, Pulse Velocity, a11d Rebound umber, '
ACJ Journal Proceedings, V. 77
o. 3, pp. 1 6 1 - 1 70.
Malhotra, V.
., and arino, . J., ed . . 2004 Jlandbook
on ondestructive Te ting ofConcrete, econd edition, CR
Press Inc . . Boca Raton, FL.
Mandel, J ., 1 984, "Fitting traight Line When Both ari­
ables are ubject to Error," Journal ofQuality Technology, .
1 6, o. I , pp. 1 - 1 4 . doi : I 0. 1 0 0/00224065 . 1 984. 1 1 97 8 1
.
•
,
(aCii
Munday, 1 . G. L., and Dhir, R. K., 1 9 4, "Assessment of In
itu Concrete Quality by Core Testing,' In Sill ondestruc­
tive Testing of Concrete, P-82, . M. Malhotra, ed. Amer­
ican oncrete I nstitute, Farmington Hill M I , pp. 393-4 1 0.
Murphy, W. E., 1 984, "The I nterpretation of Te ts on the
trength of Concrete in tructures in Situ ondestructi1•e
Testing of Concrete, P- 2, V. M . Malhotra, ed. , merican
oncrete In titute, Farmington I I i lls M I , pp. 377-392.
Murray A. M., and Long, A. E., 1 9 7 "
tudy of the
In- itu ariability of oncrete U ing the Pull-off Method,"
Proceedings - Institution of ivil Engineers, . 3, o. 4, pp.
73 1 -745. doi : 1 0. 1 6 O/i icep. l 9 7. 1 97
asser K. W., and 1-Manaseer, A. A., 1 987a "
ew
onde tructive Test,'' Concrete International V. 9,
Jan., pp. 4 1 -44.
asser, K. W., and AI-Mana eer, A. A. 1 9 7b " ompar­
i on o f onde tructive Te ters of l l ardened oncrete," A Ci
Materials Journal, . 84, o. 5, Sept.-Oct., pp. 3 74-380.
atrella, M., 1 963, Experimental Stati tics " l landbook
o. 9
ational Bureau of tandard , U . . Government
Printing Office, Wa hington, D .
ur e, R. ., 1 949, " team uring of oncrete," Maga­
zine of oncrete Research, . I , No. 2, pp. 79-88. doi :
I 0. 1 680/macr. I 949. I .2 .79
Otto en, . ., 1 9 I " onlinear Finite Element Analysis
of Pullout Test," Journal of tructural Division, V. I 07, T4
Apr., pp. 59 1 -603.
Petersen, C. G .. 1 9 4, "LOK-Test and
PO-Test Development and Their Application , ' Proceedings - Institution
of ivil Engineers, V. 76, o. 2, May, pp. 539-549. doi :
1 0. 1 6 O/i icep. l 9 4. 1 258
Peter en,
., 1 997, "LOK-TE T and APO-TEST
Pullout Te ting Twenty Year ' Experience," Proceedings of
"
•
'
the onference on Non-Destructive Testing in ivil Engi­
neering, J . H . Bungey, ed. British I nstitute of on-de true­
rive Te ting pp. 77-96.
Phoon, K. K: Wee, T. H.; and Loi, . . 1 999, "Develop­
ment of tatistical Quality A surance riterion for oncrete
ing U ltra onic Pul e Velocity Method," A '1 Materials
Journal, V. 96. No. 5, ept.- ct., pp. 568-573.
Pinto, R. C. A., and I lover, K. C., 1 999a, "E timation of
Modulu of Ela ticity in H igh-Petformance oncrete M ix
at Earlier and Later ge ," in lligh Pe1jonnance oncrete Pe1jormance & Quality ofConcrete Stru lures, P- 1 6 Amer­
ican oncrete Institute Farmington H ills, M I . pp. 5 1 -596.
Pinto, R. C. A., and Hover, K. ., 1 999b, "Application
of Maturity Approach to etting Times," ACI Materials
Journal, V. 96, No. 6, ov.-Dec., pp. 6 6-69 1 .
Popovics, J . . ; pal vier A.; and H a l l , K . ., 20 1 6, " val­
uation of P
Pavement and tructure oring and In itu
Testing Alternatives," Research Report F/IWA-'JCT- 16-022,
l l linoi
enter for Transportation, Urbana, l L .
Popovic , S., 1 998, Strength and Related Properties of
Concrete: A Quantitative Appmaclt, John Wiley & Sons,
ew York, 535 pp.
RJL M
ommis ion 42- EA, 1 9 I "Propertie of
oncrete at arly Ages- tate-of-the-Art Report," Mate-
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
rials and Structures, Research and Testing, R J LEM, Paris,
1 4,
V.
o. 84, ov-Dec, pp. 399-450.
amarin, . and Dhir R. K., 1 984, "Detern1ination of
I n itu Concrete trength: Rapidly and
onfidently by
onde tructive Testing," in if11 ondestructive Testing of
oncrete, P- 2. V. M. Malhotra ed. American Concrete
I nstitute, Farmington l l ills, M l , pp. 77-94.
Samarin, . and Meynink, P., 1 98 1 , "U e of Combined
Ultrasonic and Rebound Hammer Method for Determining
Strength of Concrete tructural Members," oncrete inter­
national, V. 3, o. 3, Mar., pp. 25-29.
Saul, A.
. 1 95 1 , "Principles nderlying the team
uring of oncrete at Atmospheric Pressure," Maga=ine
of oncrete Research, V. 2, No. 6 Mar., pp. 1 27- 1 40. doi:
I 0. 1 6 O/macr. l 95 1 .2 .6. 1 2 7
nedecor, G. W., and ochran, W. G., 1 967, tatistical
Methods, ixth edition, Iowa tate University Pre s, pp. 32-65.
Soutso , M.
.; Bungey, J. H . · and Long
. E. 2005,
"Pull out Te t Correlation and I n-Situ Strength A se sment
-The European oncrete Frame Building Project," A I
Materials .Journal, . 1 02, o. 6, ov. - Dec., pp. 422-42 .
outsos, M. .; Bungey, J. H.; Long, A. E.; and Henderson,
G. D., 2000 '·Jn- itu trength Assessment of oncrete­
The European Concrete Frame Bui lding Project,' Proceed­
in s of the 5th International Conference on DT in Civil
Engineering, T. omoto ed. Apr., !sevier cience, Tokyo,
pp. 5 3-592.
Stoll, . W.. 1 985, " ompressive trength Measurement
With the Stoll Tork Test," Concrete international, V. 7. o.
1 2, Dec., pp. 42-47.
Stone, W. . and arino, N . J., 1 9 3, "Deformation and
Failure in Large- cale Pullout Te t " A CI Journal Proceed­
ings . 0, No. 6 o .-Dec. pp. 50 1 -5 1 3 .
arino, . J ., 1 9 4 " om pari on of
tone, W.
Analytical with Experimental I nternal train Di tribution
for the Pullout Te t," ACI Journal Proceedings V. I , o. I ,
Jan.-Feb. pp. 3- 1 2.
tone, W. . · arino
. J .; and Ree e, . 1 9 6, " tali tical Method for In-Place trength Prediction by the
Pullout Test," A I Journal Proceeding , . 83, o. 5, ept.­
Oct., pp. 745-755.
Stone, \ . ., and G iza, B. J. 1 9 5, "Effe t of Geom­
etry and ggregate on the Reliability of the Pullout Te t,"
oncrete International, . 7, o. 2, Feb., pp. 27-36.
. P., 1 986, " ew tatistical
tone, W. ., and Reeve
Method for Prediction of oncrete trength from I n-Place
Test ," ement, Concrete and Aggregates, . , o. I pp.
3- 1 2. doi: 1 0. 1 520/
1 0048J
turrup, . R: Vecchio F. J.; and aratin, H., 1 9
elocity as a Measure o f Concrete om pre si e trength "
In Situ/ ondestmctive Testing of Concrete, P-82 . M .
Malhotra, ed. American
oncrete I nstitute, Farmington
H i l l , M l , pp. 20 1 -227.
wamy, R. ., and Al-Hamad, A. H. M. .. 1 984, "E alu­
ation of the Windsor Probe Test to A ses In itu oncrete
trength," Proceedings of the institution of Civil Engineers
,
43
(London), . 77, No. 2, June, pp. 1 67- 1 94. doi: 1 0. 1 6 0/
i icep. l 9 4. 1 240
Tanigawa, Y.; Baba K.; and Mori H., 1 9 4, "E timation
of oncrete trength by ombined Nonde tructive Te ting
Method, In itu ondestructive Testing of oncrere, P-82,
V. M . Malhotra ed.
merican oncrete Institute, Farm­
i ngton l l ill M l , pp. 5 7-76.
Teodoru, G. V. , 1 9 6, "Mechanical trength Property of
oncrete at Early Age a Reflected by chmidt Rebound
N umber, Ultra onic Pul e elocity, and U ltra onic ttenu­
ation," Properties of oncrere at Early Ages, P-95, J. F.
Young, ed., American Concrete Institute Farmington H i l ls,
M I , pp. 1 39- 1 53 .
Teodoru, G. V., 1 9 8 , "The s e of imultaneous ondestructive Te t to Predict the ompres ive trength of oncrete "
ondestructive Testing, P- 1 1 2, H . . Lew., ed., American
oncrete Institute, Farmington Hill , M I . , pp. 1 37- 1 52.
Tesfamariam, S.; Sadri, A.; and Thoma , M . 2006,
•-combined Stre Wave Technique to Estimate Con rete
trength," Material Evaluation, . 64, o. 1 2, pp. 1 1 73- 1 1 79.
Tex-426- , 20 1 0 "E timating oncrete trength by the
Maturity Method," Texas Department of Tran portation,
u tin, TX.
ay burd A. M . , and Mc Donald, J. E., 1 999, " n valu­
ation of Equipment and Procedures for Tensi le Bond Testing
of oncrete Repair " Technical Report REMB- - 6 1
rmy orp of Engineers, Waterway
xperiment tation,
June, p. 4.
incentsen, L. J ., and Henriksen, K. R., 1 992, "Denmark
Span Strait with Great Belt Link," Concrete International,
. 1 4, No. 7, July, pp. 25-29.
ogt, W. L.; Beizai : and Dilly, R. L. 1 984, ' I n itu
Pullout
trength of Concrete with Insert Embedded
by ' Finger Placing'," In Situ ondestructive Testing of
oncrete, P- 2, . M . Malhotra ed., American oncrete
I nstitute Farmington Hill , M l , pp. 1 6 1 - 1 75.
Wade, . A: ixon J . M : chindler, . K . : and Barnes,
R.
., 20 I 0 " ffect of Temperature on the
etting
Behavior of oncrete," .Journal of Material in ivil Engi­
neering, V. 22, o. 3, Mar., pp. 2 1 4-222. doi: I 0. 1 06 1 /
(ASCE)O 99- 1 56 1 (20 1 0)22 :3{2 1 4)
Whiting, D: agi, M.; Okamoto, P.; and Delaney, H . ,
1 994, "Quality ontrol of oncrete o n - ite-U e r ' Manual,"
Strategic 1/ig/rways Research Program Report (SIIRP- 4 1 4), ational Research Council, Washington, D , 5 1 pp.
Yener, M., 1 994, "Overview and Progressive Finite
Element Analysis of Pullout Tests," A I Structural .Journal,
. 9 1 , No. I, Jan.-Feb. pp. 49-58.
Yener, M., and hen, . F., 1 984, ''On In-Place trength of
oncrete, and Pullout Te ts, Cement, Concrete and Aggre­
gates, V. 6, o. 2, pp. 90-99. doi: 1 0. 1 520/CCA 1 036 1 J
Yun, . I I .; Choi, K . R.; Kim, S . Y.; and ong, Y. . , 1 9
'·Comparative Evaluation o f onde tructive Test Method
for In-Place
trength Detennination,"
ondestructive
Testing, P- 1 1 2, H . S. Lew, ed., American oncrete I nsti­
tute Fatmington H i lls, M l , pp. 1 1 1 - 1 36.
'
American Concrete Institute - Copyrighted © Material - www.concrete.org
44
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
10 -.---,:--�----,
APPENDIX A
A.1-Minimum nu mber of strength levels
The minimum number of trength levels required to
develop the strength relation hip depends on tatistical
con ideration and cost. To gain ome insight it is useful
to examine how the confidence interval for an e t imate
obtained from a strength relationship is affected by the
nwnb r of point u ed to e tablish that relation hip (Carino
1 993 ). Because the trength relation hip i u ed to e timate
compressive strength from in-place te t results, compressive
strength i treated as the dependent variable ( Y value) and
the in-place result as the independent variable (X value).
The residual standard deviation, also identified as standard
error of e timate, i the basic parameter used to quantify the
uncertainty of a best-fit strength relationship for a given set
of data. For a linear relation hip, an estimate of the re idual
standard de iation is as fol lows
s
•
=�:r.(d,.Y
-2
(A. l a)
where S, is estimated residual standard deviation; d,x is devi­
ation of each test point from the best-fit l ine; and is number
of te t points u ed to establish the trength relationship.
When the strength relationship i used to e timate the
mean value of Y at a new value of X, the width of the confi­
dence interval for the mean is related to the residual stan­
dard deviation by the fol lowing expression ( atrella 1 963 :
Snedecor and Co hran 1 967)
{ A. I b)
where W is width of the I 00( 1 - a) percent confidence
interval for the estimated mean value of Y for the value X
t,v.2.a.1 is tudent's /-value for -2 degree of freedom and
significance le el a; X i the average of X values u ed to
develop strength relation hip; and S,,_, is um of quare
of deviation about X of the X values used to develop the
strength relationship Sxx - L(X X )2 •
trietly peaking, Eq. (A. I b) i applicable 011ly if the
a umption of ordinary lea 1- quare analy is are atisfied.
It is used here to demon trate, in a implified way, the effect
of the nwnber of test points on the width of the confidence
interval. In practice, Eq. (A.3) hould be used to determine
the lower confidence limit of the estimated mean value of Y
for a new value of X.
The econd term under the square root sign in Eq. (A. I a)
show that the width of the confidence interval increase a
the distance between X and X increases. This means that
the uncertainty of the e timated trength is greater at the
extreme l imits of the trength relationship than at its center.
To examine how the width of the confidence interval i
affected by the number of test point , con ider the case where
X = X o that the second term under the square root sign in
ro
2:
2 W(X)
E
�
c
Q)
"'0
1;::
c
0
()
-
S
e
5-
.
. . .
.
.. .
• .
. .. .
Recommended
Range
.
. . .
l"Lt 1
· ·
:
�'
0
'
'
:
· · · · · · · · . . · · · · ·
.
.
. .
:
0
0
:
. . . . . . . . . . . . .
..
:
0
0
0
0
• • •
:
:
!
0
0
• • • •
0 4-���-�.���+.-����+,·��
20
15
0
10
5
N u mber of Test Points, n
Fig. A. I-Effect of number of point u ed to establish
strength relation hip on the confidence interval width (in
terms of residual tandard deviation).
q. (A. l b) equal zero. The width of the confidence interval
relative to the rc idual tandard deviation i as follows
W(
X
S
,
(I
) = 2tJ\ -2.al2 'JN
(A. l c )
Equati n ( . I c) i plotted i n Fig. A . I t o how how the idth
of the 95 percent confidence interval (relati e to S,) i affected
by the number of tc 1 poin u cd to e tabli h the trength
relation hip. It i ccn that for fc\ tc t point ( ay lc than
5) by in luding an additional te t p int there i a ignificant
reduction in the relati e width of the confidence interval.
For many point , however the reduction obtained by u ing
an additional te I point i smal l. Therefore, the appropriate
number of strength levels is determined by con ideration of
preci ion and co t. It has to be decided if it i worth the addi­
tional expen e to obtain additional preci ion by u ing another
te t point. From Fig. . I it i rea onable to conclude that the
minimum number o f te t point i about i , v hile more than
nine tc 1 would probably not be ju tificd economically.
A.2-Regression analysis with X-error (Mandel
1 984)
I f th procedure in 7.2.3 or 7.2 .4 are to be u ed to e timate
the in-place characteri tic trength, the lea t- quare regre ion analy i procedure to determine the trength relation­
ship should account for error in the X- ariable. The method
propo ed by Mandel ( 1 9 4) can be u ed for thi purpo e.
Thi section pro ide a tep-by-step procedure for carrying
out the Mandel method.
At each trcngth level for the correlation te t there arc n,
replicate in-place tc t re uh and n_. replicate omprc i c
te t rc uh . The number of trength le el i
. The objec­
tive is to find the be t-fit alue of a and 8 (and their uncer­
taintie ) for the traight line, trength relation hip
In
= a + 8 In!
American Concrete Institute - Copyrighted © Material - www.concrete.org
(A.2a)
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228.1 R-19)
where a is the intercept of straight line; B is the slope of straight
line· InC i the natural logarithm of com pre sive strength; and
In/ is the natural logarithm of the in-place test result.
After the correlation te t data have been obtained, the
fol lowing sequence of calculations is u ed to establish the
trength relation hip and its uncertainty:
I . Tran form the data by taking the natural logarithm of
each te t re ult
x - lni
(A.2b(a))
y = Inc
(A.2b(b))
where i and c are the individual in-place and compre i e
trength test results, respectively.
2. For each strength level j, compute the average and
standard deviation of the logarithms of the in-place and
compres ive te t result :
0 = the average of the logarithm of the in-place te ts at
strength level j
>j = the average of the logarithms of the compre si e
strength tests at trength levelj
the standard deviation of the logarithms of the
s_.,
in-place test at trength levelj
s,i - the tandard de iation of the logarithms of the
compressive strength test at strength levelj
3 . a leu late (sx)1 and (s>)2, which are the average variances
(squares of the standard deviations) of the logarithm of the
in-place tests and of the compressive te ts re pectively.
(
)
)
2
E ( s f )l
· =-
5. Find the values of b and k by sol ing the following
simultaneous equations
(A.2e(a))
k = !!...
A.
(
n,
\ here llx is the number of replicate in-place test
at each
strength level; and n, is the number of replicate com pres i e
strength test at each strength level.
The numerator and denominator in q. (A.2d) are the
variances of the average compres ive strength and in-place
result , respectively. If there are different numbers of replicate
test at each trength level the a erage numbers of replica­
tion hould be u ed for "·• and n_, ( tone and Reeve 1 9 6).
.2f(a) )
(A.2f(b))
(
.2f(c) )
The terms X and Y are the grand averages of the loga­
rithm of the in-place and compressive treng1h te t results.
LX
'
X = --
(A.2c{b))
(A.2d)
(A.2e(b))
An iterative procedure can be used to solve for k and b
( Mandel 1 9 4). First, assume a value of k such as k = 0, and
ol e for b in Eq. (A.2e(a)). Using this value of b olve for a
new value of k in Eq. ( .2e(b)). ubstitute the new value of k
into q. (A.2e(a)) and solve for b. Repeat the procedure until
the values of k and b converge, which will usually occur in
les than five iteration .
In Eq. {A.2e(a)), the term S.u. s_,,, and S.ry are calculated
according to the fol lowing
(A.2c(a))
quations (A.2c(a)) and ( .2c(b)) a ume that the same
number of replicate tests are used at each trength levei. I f orne
te t re ult have to be discarded because th y are outliers, the
pooled variance hould be computed to accow1t for different
numbers of replicate te t at each trengtb level (refer to tone
and Reeve [ 1 986] or a textbook on introductory statistics).
4. ompute the value on.. a follow
45
- EY
Y='
(A.2g(a))
(A.2g(b))
6. The be t-fit e timate of B and a are as follow
B-b
(A.2h(a))
a = Y - bX
(A.2h(b))
7. Use the fol lowing steps to compute the tandard errors
of the estimates of a and B.
(a) ompute these modified sums of quares
.,,
=
"' + 2k .n + �s,�
(
.2i(a))
(
.2i(b))
(b) Compute the following error of fit, s,
s.
=
(c) The error in a is gi en by the following
American Concrete Institute - Copyrighted © Material - www.concrete.org
(A.2j )
46
REPORT ON METHODS FOR E STIMATING IN-PLACE CONCRETE STRENGTH (AC1 228.1 R-19)
s.
=
s.
(A.2k)
where S! is the standard de iation of estimated alue of Y
(average concrete strength);
is number of point used to
obtain the strength relationship; h i e timated slope of the
strength relation hip; k b!A., where A. is obtained from the
single-operator variability during correlation te ting (Eq.
( .2d))· X is the average of the logarithms of the in-place
tests performed on the structure; X i grand average of X
alues during correlation te ts (Eq. (A.2g(a))); , i error of
fit of treogth relation hip ( Eq. ( .2j )): Suu i the modified
sum of the quare a gi en by Eq. (A.2i(a))· sx i tandard
deviation of the logarithms of the in-place test performed
on the structure; and m is the number of replicate in-place
test done on the tructure.
Equation (A.3) shows that there are two sources of the
uncertainty in the e timated value of Y:
( I ) The uncet1ainty of the trength relationship (s,)
(2) The uncertainty of the in-place test result obtained
from testing the structure (s..\)
Because Eq. ( . 3 ) i the sum of two variances, which
could have d ifferent degree of freedom, a fonnula ha been
sugge ted for computing the effective degrees of freedom
for sr ( tone and Reeve 1 986). For simplicity, it can be
a sumed that there are (m- 1 ) degree of freedom as ociated
with St, where m is the number of in-place tests done on the
structure. These degree o f freedom are used in choosing the
t-value to calculate a lower confidence limit for the average
alue, as di cu sed in 7.2.4.
=
(d) The error in B is given by the fol lowing
a=
{A.21)
In ummary, the fol lowing general tep are u ed to obtain
the best-fit strength relation hip and account for the error in
the X variable ( in-place test re ult ):
( I ) Transform the correlation data by taking their natural
logarithms
(2) At each strength level, compute the average and stan­
dard deviation of the tran formed values ( logarithms)
( 3 ) ompute the value of A. based on the average (or
pooled) ariance of the mean compres ive and in-place
results
(4) ompute the value of b and k
(5)
ompute the slope and intercept of the be t-fit
relation hip
(6) ompute the error of the fit
The error of the fit, s., is needed to calculate the uncer­
tainty in the e t imated mean compressi e strength when the
strength relationship is used with in-place tests of the struc­
ture. This is explained in the next section.
A.4-Example
A.J-Standard deviation of estimated Y-value
(Stone and Reeve 1 986)
The strength relation hip is used to estimate the in-place
compre sive strength ba ed on the re ult of the in-place
te t done on the tructure. Typical ly, everal in-pia e te t
are done on the structw-e, the average re ult i computed,
and the strength relationship is used to estimate the average
compressive strength. To obtain a reliable estimate of the
a erage strength-that is a value that has a high probabi l ity
of being exceeded-the standard deviation of the estimate
must be known.
The approach developed by Ylandel ( 1 9 4) can be u ed
to e timate the standard deviation of an estimated value of Y
(average compres ive strength) for a ne'V value ofX (average
in-place te t re ult ) when there i X-error. Mande l ' method
wa modified by tone and R C\'e ( 1 9 6) o that it al o incor­
porate the uncertainty of the average in-place result from
tests on the structure. This modification account for the fact
that the uncertainty in the average of the in-place results is
typically greater for te ts on the structure compared with
that from the laboratory tests used to develop the trength
relation hip. The tandard deviation of the e timated value
of Y (average of the logarithm of compre sive trength) i
obtained by the following equation
2
z� + bz _
.
111
(A.3)
An example is presented to how the application of the
Mandel { 1 9 4) method and illu trate the e aluation of
in-place te t u ing the tolerance factor method di cu ed
in 7.2.2 and the alternati e method di cus ed in 7.2.4. The
correlation data are taken from the study of the pullout test
by tone et al. ( 1 9 6 ). The pullout test geometry had an
apex angle of 70 degrees and the concrete was made u ing
river-gravel aggregate. ight strength level were u ed to
develop the strength relationship. At each strength level, I I
replicate pullout te t and five replicate cyl inder compre sive te ts were done.
soft conver ion of the inch-pound
alue reported by Stone et al. ( 1 9 6) � as used to obtain the
corre ponding I value .
The data from the cited reference were converted by
taking the natural logarithm of the individual pul lout loads
and compressive strengths. The average standard devia­
tion, and variance (square of tandard deviation) of the
transformed pul lout load at each strength level are hown
in olumns I , 3 , and 4 of Table .4a. ( I and inch-pound
er ion of orne tables are pre ented in this ppendix to
reduce clutter.) The a erage, standard deviation, and vari­
ance of the transformed compres ive strength at each
trength level are hown in olumn 5 7 and . For infor­
mation Column 2 and 6 gi e the a erage of the loga­
rithm value tran formed into real units.
The average values in olwnns I and 5 of Table .4a
were used to calculate the various parameter to establish
the strength relationship according to the procedure in .2.
American Concrete Institute - Copyrighted © Material - www.concrete.org
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (ACI 228. 1 R- 1 9)
47
Table A.4a-Average, standard deviation, and variance of correlations data from Stone et al. (1 986)
Average lnPO
for PO. k (lb)
1
Real value of
PO, kN (lb)
tandard
deviation lnPO
I
2
3
2.26 9 (7 .6842)
9.67 (2 1 74)
0. 1 085
(7.9 1 3 )
1 2. 1 6 (273 )
0.0459
2. 076 ( .22-9)
1 6.57 (3725)
0.0700
2.988
( .4().10)
1 9.36 (4465)
0. 1 065
3.2945 (8.7098)
26.97 (6062)
0. 1 1 62
3.3948 (8. 8 1 00)
29. 8 1 (670 I )
3. 5244 (8.9397)
3.5725 ( 8.9877)
2.49
-
-
ariance
InPO
4
n,.
x
y
-
ariance In
tandard
deviation InC
5
6
7
8
I 0.39 ( 1 508)
0.0474
0.0022
0.002 1
2.6522 (7.6292)
1 4. 1 9 (2057)
0.043
0.0049
2.9273 ( 7.9().13)
I .68 (2709)
0.0 1 1 4
3. 1 275 ( . 1 047)
22.82 ( 3 3 1 0)
0.0 1 35
3.3440 (8.3209)
28.33 (4 1 09)
0.0343
0. 1 488
0.0222
3.45 5 1 (8.432 1 )
3 1 .66 (4592)
0.0048
0.0005
33.93 (7629)
0.0953
0.009 1
3.6890 c .6660)
40.00 (5802)
0.507
0.0026
35.60 ( 8004)
0. 1 598
0.0255
3.7588 (8.7358)
42.90 (6222)
0.0303
0.0009
-
-
Average variance of lnPO
0.0 125
Value. I umts
(in.-lb units)
Parameter
k
Value. I unitS
(in.-lb units)
4.287 (4.2 4 )
11 (11)
h=B
1 .0 0 ( 1 .030)
5 (5)
{I
0.0268 ( 0.5747)
3.0438 ( 8.4590)
I'"
0.240 (0.240)
s...
1 .027 (0.563)
1
I
3.6 -r0>
c
E
-r..s:::. � 3.4
� U5 3.2 -rtV Q)
C) - >
0
_J
tV
in
[;
......
2
tV E
0
2.6 -1-
0
2.4 -
z u
'+-
r- .
2.2
2.2
SB
0.0001
=
0.00 1 4
0 0268 + 1. 030 /n(PO)
./
/
I
I
-
0.00 1 2
.
./
/
I
I
/
I
•
I
3.4
2.4
3
3.2
2.6
2.8
Natural Logarithm of Pullout Load
/
•
3.6
8.8 .....----�---,
A computer preadsheet wa et up to do the e calculation .
Table .4b ummarize the calculated alues.
The calculated values of a and 8 are hown in the Ia t
column ofTable A.4b. Therefore, the equation of the trength
r lation hip is a follow
I nch-pound unit : In
/n(CS)
-r-
...... a. 2.8
-
S,.
I units: InC = 0.026 + 1 .0301nPO
-
0.0020
-
0.0103
(a )
3 -i-
s.
Sn.
0.045 1
0.00 1 9
3.8
..s:::.
s,
s,
r-
-
Average variance of InC
4 . 1 55 (4 . 1 04)
. 1 6 1 9 ( . 1 3 9)
I.
Real value of C.
MPn (ps1)
2.34 1 3 (7.3 1 3)
8 (8)
"·
Average InC
for , MPa (psi )
0.0 1 1
Table A.4b-Summary of resu lts of regression
calculations using values in Table A.4a and
procedure in Appendix A.2
Parameter
1
(A.4a)
= -0.5747 + I .0301nPO(A.4b
where InC is the average of natural logarithms of compre sive strengths; and lnPO i average of natural logarithms of
pullout load .
F igure A.4 shows the correlation data (a erage of loga­
rithms) and the best-fit line.
Finally the trength relation. hip and the procedures in
7.2 arc u ed to e timate the in-place compre ive trength
based on in-place te t re ult . Table .4c how two et of
in-place pullout test re ult . Both ca e have approximately
the arne a erage alue, but a e 2 ha higher ariability.
In each case, there are I 0 replicate test results-that is, 111 =
I 0. The pullout loads are transfonned by taking their natural
logarithms. The averages of the logarithms, lnPO, are substi­
tuted into Eq. ( .4) to obtain the average of the logarithm f
..s:::.
0>
8.6
·;:::: C/)
Q)
tV
8.2
_J
8
(b)
'"iCS)
E c 84
�
..s:::.
- -
•
-o 5747 • 1 030 /"(PO
•
g> ,�
en
[;
� a.
2 E
Z
(V 0
(.)
•
•
./
./
7.8
7.6
0 7.4
/
•
.L...J L...L.'-+
. ..._._-'-+.......J.
..
....L.
...f ....I.
..I. -'...j L..I.f-'-J...
...
..I...-i
7.2 -+-'....'-+
7.6
7.8
8
8.2
8.4
8.6
8.8
Natural Logarithm of P u llout Load
9
Fig. A.4-Datafor strength relationship and best-fit line: (a)
SJ units; and (b) inch-pound units.
in-place compressi e strength, In . Estimates of the tenth
percentile strength ( Y0_ 10) corresponding to the two cases are
obtained using the tolerance factor method ( 7.2.2 ) and the
altemati e method ( 7.2.4). The values of the various param­
eters used in the calculation are ummarized in Table A.4d
and ' here appropriate, the corresponding equation number
are hown. For the altemative method the tandard dcvia-
American Concrete Institute - Copyrighted © Material - www.concrete.org
48
REPORT ON METHODS FOR ESTIMATING IN-PLACE CONCRETE STRENGTH (AC1 228. 1 R· 1 9)
Table A.4c-Val ues of pul lout force obtained from tests on structures
In inch-pound rmifs:
e I
Pullout force, ·
LnPO
a c2
LnPO
Pullout force. lb
2. 944
1 7. 7
2. 545
30 1 0
3904
2.69 5
1 2.78
2.5479
3340
2 73
2.7453
1 4.25
2.6569
3204
2.6 1 74
II. 7
2.4742
2669
2.4000
10.37
2.3392
2332
2.59 1 1
1 3.75
3000
3091
2.6834
1 7. 1 0
3290
3844
2.6 1 42
1 3.97
3070
3 1 40
1 1 .8
2660
2552
1 1 .83
2660
3336
Average (.A')
Average (X)
1 3.39
A\'eragc (X)
Average (X)
tandard deviation ( x)
lnPO
LnPO
Pullout force, lb
0. 1 670
tandard d viation (s.,)
Table A.4d-Estimate of i n-place com pressive strength usi ng results i n Table A.4c
Alternative app roac h (7.2.4)
1--f-
1-
Y (Eq. (A- 1 7a))
c p( Y), MPa•
1-
1-
Sy
-
r-
(Eq. (A.3))
IQ,OOS
Y1.,. (Eq. (6-4))
e I
1 4.7
0.0454
-
i�
--
--
1 -. l o
0.0607
I . 33
1 .833
2 .6098
2.6034
0.037
0.055
exp( Yo rol ( Eq. (6-5))
2.5628
2.5326
1 2.97
1 2.59
-
r-
Y (Eq. ( A- 1 7b))
cxp( Y), p i*
St
(Eq. (A.3))
IQ.O Ol
r-
-
(7 .2.4d))
cxp( Yo IO) (Eq. (7.2.4c))
e�tp( Yo r o), p i
'cxp( l') - e'.
Alternative approach (7.2.4)
Case I
7.6700
2 1 43
Y,.,. ( Eq. (7.2.4b))
Srf (Eq.
ase I
!-
-
,_
y
2.6930
expO'). MPa
1 4.78
K (p = 0.75)
-r-
s, ,
I
Yo r o (Eq. (6- 1 ))
exp( Yo w), M Pa
In in.-lh rmi1s:
1-
I-
Tolerance factor approach (7.2.2)
ase 2
2.7 1 47
srr ( Eq. (7.2.4d))
exp( Yo ro). MPa
r
I
a
2.6930
In Sl rmifs:
-
--
-f-
ase 2
7.69 1 7
2 1 90
0.0454
0.0607
1 .833
1 .833
7.5870
7.5804
0.037
0.055
7.5395
7.5099
18 I
1 826
r-
-
!--
-
tion of the in-place compres ive strength (sd) was computed
u ing q. (7.2 . 4d), while for the tolerance factor method it
wa taken to equal the tandard deviation of the tran formed
in-place tc t re ult . For each method, the alue of Y0. 1 0 i
a smaller fraction of the average strength for asc 2 due to
II
y
1 .67 1
i-
�
1-
-
- --
a
e2
2.7 1 47
1 5. 1 0
1 .67 1
0. 1 1 1
0. 1 67
2.5075
2.4356
1 2.27
1 1 .42
Tolerance fact r nppr ach (7.2.2.)
ase I
7.6700
--
a
e2
7.69 1 7
cxp( Y). psi
2 1 43
2 1 90
K (p = 0.75)
1 .67 1
1 .67 1
s, ,
0. 1 1 1
0. 1 67
7.4845
7.4 1 26
1 780
1 657
Yo 111 (Eq. (7 .2.2a))
exp( Y0 10), psi
-
-
-
-
-
-
-
-
the higher variability of the in-place te ts. In this example,
the trength relationship ha relatively low catter, and the
estimates of Y0.10 are lower for the tolerance factor method,
which does not consider this.
American Concrete Institute - Copyrighted © Material - www.concrete.org
American Concrete Institute
Always advancing
As ACI begins its second century of advancing concrete knowledge, its original chartered purpose
remains "to provide a comradeship in finding the best ways to do concrete work of all kinds and in
spreading knowledge. ' In keeping with this purpose, ACI supports the following activities:
·
Technical committees that produce consensus reports, guides, specifications and codes.
·
Spring and fall conventions to facilitate the work of its committees.
·
Educational seminars that disseminate reliable information on concrete.
·
Certification programs for personnel employed within the concrete industry.
·
Student programs such as scholarships, internships, and competitions.
·
Sponsoring and co-sponsoring international conferences and symposia.
·
Formal coordination with several international concrete related societies.
·
Periodicals: the ACI Structural Journal, Materials Journal, and Concrete International.
Benefits of membership include a subscription to Concrete International and to an ACI Journal. ACI
members receive discounts of up to 40% on all ACI products and services, including documents, seminars
and convention registration fees.
As a member of ACI, you join thousands of practitioners and professionals worldwide who share
a commitment to maintain the highest industry standards for concrete technology, construction,
and practices. In addition, ACI chapters provide opportunities for interaction of professionals and
practitioners at a local level to discuss and share concrete knowledge and fellowship.
American Concrete Institute
38800 Country Club Drive
Farmington Hills, MI 48331
Phone: +1.248.848.3700
Fax:
+1.248.848.3701
www.concrete.org
American Concrete Institute
Always advancing
38800 Country Club Drive
Farmington Hills, Ml 48331 USA
+1.248.848.3700
'W
/.CQnC
Q
The American Con crete I nstitute (ACI) is a leading authority and resource
worldwide for the development and d i stribution of consensus-based
sta n d a rds and techn1cal resources, educational prog rams, and certifications
for individuals and organizations involved in conc rete design, construction,
and materials, who share a commitment to pursuing the best use of concrete.
I ndividuals interested i n the activities of ACI are encouraged to explore the
ACI website for me mbersh i p opportun ities, comm ittee activities, and a wide
variety of concrete resources. As a vol unteer mem ber-driven orga n i zation,
ACI invites partnerships and welcomes all concrete p rofess ionals who wish to
be part of a respected, connected, socia l group that provides an opportun ity
for professional growth, networking and e njoyme nt.
. I��JI�II II Wl�J] I