NATECH2016
January 12-13, 2016
Nakanoshima Center of Osaka University
SEISMIC RESISTANCE
CAPACITY ON PIPE BRACED
SUPPORTING FRAME OF
SPHERICAL TANK
Takashi OHNO
The High Pressure Gas Safety
Institute of Japan (Graduate
School of Aoyama Gakuin
University)
Nobuyuki KOBAYASHI
Aoyama Gakuin University
Kenji OYAMADA
The High Pressure Gas
Safety Institute of Japan
OUTLINE
1. Introduction
 seismic damage for high pressure gas equipment in
japan in 2011
 Investigation into the cause of accident
 objective ～ evaluation of seismic reinforcement～
2. Stress analysis for Spherical tank using FEA
 Deformation behavior
 Stress distribution
3. Evaluation of seismic capacity using FEA
 Plastic strain distribution
 Seismic capacity
4. Evaluation of seismic capacity using Shaking tests
 Experimental conditions
 Result of response character and seismic capacity ,
fracture mode from Shaking test
5. Summary
INTRODUCTION(1/3)
Huge earthquake hit East Japan (March.11th.2011)
→Fire and explosion accident (LPG spherical tank was collapsed )
→Investigation into the cause of accident
the 2011 Great East Japan Earthquake
Epicenter of Main shock
date and time ： 2011/3/11 14:46:18
magnitude ： 9.0
××
Epicenter of Max. After shock
date and time ： 2011/3/11 15:15:34
magnitude ： 7.6
Accident 2 :Spherical tank support broke
Accident 1:Spherical tank collapsed
INTRODUCTION(2/3)
Huge earthquake occurred in Japan (March.11th.2011)
→Fire and explosion accident (LPG spherical tank was collapsed )
→Investigation into the cause of accident
Accident 1
Upper column
Lower column
Fig. Spherical Tank with steel pipe
brace and column structure
Collapsed spherical tank
Steel pipe brace breaking
Accident 2
INTRODUCTION(3/3)
Huge earthquake occurred in Japan (March.11th.2011)
→Fire and explosion accident (LPG spherical tank was collapsed )
→Spherical tank support was broken
→Investigation into the cause of accident
Main shock (14:46)
Intersection of braces breaking
after shock (15:15)
Columns buckling and tank collapse
accident causation ： strength poverty of intersection of braces
→need reinforcing intersection of braces
Objectives of this study
Validation:Fracture mechanism and deformation behavior
Study:Effectiveness of reinforcement and seismic capacity of spherical tank
ANALYSIS MODEL OF SPHERICAL STORAGE
Intersection of braces
TANK FOR FEA
Outer surface
Cross-section
Case1
None
Upper
column
Short
brace
Long
brace
Lower
column
Case No.
Case 1
Case 2
Case 3
Brace
without reinforce
with diaphragm
with gusset plate
Column
without reinforce
with diaphragm
with diaphragm
Number of nodes
Number of elements
936,447
313,785
757,064
254,734
774,693
261,433
Case2
Diaphragm
Case3
Gusset plate
ANALYSIS CONDITION FOR FEA
Plane of structure
Max compression to LB
Elastic analysis
Upper
column
Plane of structure
Max tension to LB
Young’s
modulus
E
Poisson’s
ratio
ν
Long
brace
Fixed Lower edge
Lower
column
Spherical
shell
205 GPa
0.3
Elasto-plastic analysis
Upper
column
Short
brace
Lower
column
Brace
Yield
stress
σys
Ultimate
Tensile
strength
σuts
S–S
Curve
model
Upper
column
Lower
column
Brace
470MPa
235MPa
610MPa
(εP=0. 2)
400MPa
(εP=0. 3)
2-liner
2-liner
Spherical
shell
―
(Elastic)
DEFORMATION OF INTERSECTION OF BRACES
BY ELASTIC ANALYSIS
Displacement (mm)
Seismic force direction
Short brace
Long brace
0.25
Long brace
Evaluating line
0, 1
0.75
0.5
Display magnification: x25
Short brace
Display magnification: x50
τ
Shear force
Long brace
Weld
Display magnification: x25
σ’t
-σ’c
Compressive
stress -σ’c
Tensile
stress σ’t
σt
τ
Weld
σ’t
-σ’c
Short brace
Shear force
τ
τ
σt
Compressive
stress -σ’c
Tensile
stress σ’t
STRESS EVALUATION OF INTERSECTION OF BRACES
BY ELASTIC ANALYSIS CASE1 (UNREINFORCEMENT)
B
A
E
D
C
FMH=8114 kN
1800
150
A:Long brace (Upper side)
100
50
1600
Cross section
E
1400
internal
1200
middle
Equivalent of Mises stress σeq, MPa
Longitudinal Stress
σL, MPa
200
D:Long brace (Lower side)
external
1000
0
B:Short brace (Upper side)
-50
C:Short brace (Lower side)
-100
-150
-200
800
600
400
200
0
0
0.2
0.4
0.6
Normalized length
0.8
1
0
0.25
0.5
Normalized length
0.75
1
STRESS EVALUATION OF INTERSECTION OF BRACES
BY ELASTIC ANALYSIS CASE2 (DIAPHRAGM)
B
A
E
D
C
FMH=8114 kN
1800
150
Equivalent of Mises stress σeq, MPa
Longitudinal Stress
σL, MPa
200
100
50
D:Long brace (Lower side)
B:Short brace (Upper side)
C:Short brace (Lower side)
-100
1400
internal
1200
middle
external
1000
0
-50
Cross section
E
1600
A:Long brace (Upper side)
-150
-200
800
600
400
200
0
0
0.2
0.4
0.6
Normalized length
0.8
1
0
0.25
0.5
Normalized length
0.75
1
STRESS EVALUATION OF INTERSECTION OF BRACES
BY ELASTIC ANALYSIS CASE3 (GUSSET PLATE)
B
A
E
D
C
200
1800
150
1600
Cross section
E
1400
internal
Equivalent of Mises stress σeq, MPa
Longitudinal Stress σL, MPa
FMH=8114 kN
A:Long brace (Upper side)
100
middle
1200
50
D:Long brace (Lower side)
0
-50
B:Short brace (Upper side)
C:Short brace (Lower side)
-100
-150
-200
0
0.2
0.4
0.6
Normalized length
external
1000
0.8
1
800
600
400
200
0
0
0.25
0.5
Normalized length
0.75
1
SUMMARY OF ELASTIC ANALYSIS ON EFFECTIVE OF
REINFORCEMENT
Case1 unreinforcement
internal
1200
middle
external
1000
800
600
400
200
Cross section
E
1400
internal
1200
middle
0
Mises stress
(MPa)
0.25
0.5
Normalized length
Long brace
0.75
1
Short brace
middle
1200
external
1000
800
600
400
200
800
600
400
200
0
0
0
internal
1400
external
1000
Cross section
E
1600
Equivalent of Mises stress σeq, MPa
1400
Case3 gusset plate
1800
1600
Equivalent of Mises stress
Cross section
E
Equivalent of Mises stress σeq, MPa
1600
Case2 diaphragm
1800
σeq, MPa
1800
0
0.25
0.5
Normalized length
Long brace
0.75
1
Short brace
0
0.25
0.5
Normalized length
Long brace
0,1
0,1
Evaluating line
Reinforced type
without reinforce
with diaphragm
with gusset plate
Short brace
Short brace
0,1
Evaluating line
Evaluating line
Short brace
0.75
Short brace
Inner surface
Middle
Outer surface
average
max
average
max
average
max
308
1624
167
619
247
1633
141
368
108
215
118
610
145
526
98
249
122
423
Long brace
σt=100 MPa
Short brace
σc=130 MPa
1
EVALUATION OF ULTIMATE STRENGTH BY ELASTOPLASTIC ANALYSIS
Case1 Distribution of equivalent of plastic strain
without reinforcement
Equivalent of plastic strain
Seismic Inertia Force
FMH=12620 kN
Maximum
plastic strain
Maximum
plastic strain
EVALUATION OF ULTIMATE STRENGTH BY ELASTOPLASTIC ANALYSIS
Case2 Distribution of equivalent of plastic strain
with diaphragm
Equivalent of plastic strain
Seismic Inertia Force
FMH=15658 kN
Maximum
plastic
strain
EVALUATION OF ULTIMATE STRENGTH BY ELASTOPLASTIC ANALYSIS
Case3 Distribution of equivalent of plastic strain
with gusset plate
TOP VIEW
Equivalent of plastic strain
Seismic Inertia Force
FMH=17531 kN
Maximum
plastic strain
Deformation scale: x20
Large plastic strain
SUMMARY OF ELASTIO-PLASTIC ANALYSIS ON
EFFECTIVE OF REINFORCEMENT
Horizontal seismic force FMH, kN
20000
Gusset plate
Diaphragm
18000
16000
14000
Unreinforcement
12000
10000
•
Case3:Seismic horizontal Force was
not constantly increasing at
calculating limit.
•
In Case3 it was considered that
braces reinforced by gusset plate
was buckled, 18,000kN is brace
limit buckling load.
8000
6000
Case 1 without reinforcement
4000
Case 2 reinforced with diaphragm
2000
Case 3 reinforced with gusset plate
0
0
20
40
60
80
100
120
Horizontal Displacement DH, mm
Fig. Relationship of Force and displacement
Case
Maximum horizontal
seismic force kN
Effectiveness of
reinforcement
Case1 (unreinforcement)
12620
-
Case2 (diaphragm)
15658
1.25 times
Case3 (gusset plate)
17531
1.39 times
SHAKING TESTS
Test models specs
Dimensional ratio for real equipment 1/1.896
Tank A：Unreinforced model
Tank B：Reinforcement model with gusset plate
same as FEA model
Displacement meter
Acceleration meter for response
Input seismic wave
・ＪＭＡ Kobe ＮＳ wave
・Simulated seismic wave
・Sine wave
Simulated seismic wave
Created by design spectrum
800
Acceleration meter for input
・Acceleration. and duration time with similarity rule
・Cut off long term period with high pass filter(2.8Hzz)
・Simulated seismic wave fit on acceleration spectrum for design
Acceleration α, gal
600
400
200
0
-200
-400
-600
-800
0
20
40
60
80
Time t, sec
100
120
TEST CASES
Similarity rule
Acceleration: original ×1.896
Time
: original / 1.896
Without reinforce
JMA Kobe NS
Reinforced with gusset plate
Tank A
Acceleration
amplification ratio β
Tank B
Maximun Input
Acceleration α imax (gal)
10%
15%
100%
Acceleration
amplification ratio β
158
227
1062
Maximun Input
Acceleration α imax (gal)
25%
50%
100%
340
745
1065
Simulated seismic wave
Without reinforce
Tank A
Reinforced with gusset plate
Tank B
Maximun Input
Maximun Input
Acceleration
Acceleration
Wave shape
amplification ratio β Acceleration α imax (gal)
amplification ratio β Acceleration α imax (gal)
5%
55
10%
103
10%
107
20%
201
15%
158
35%
335
Simulated
30%
295 Simulated
70%
693
Sismic Wave
70%
699 Sismic Wave
100%
981
100%
986
138%
1362
138%
1343
200%
2036
200%
2008
260%
2648
sin wave (11.2Hz 200gal)
175
sin wave (11.0Hz 800gal)
800
Wave shape
MOVIE DURING SHAKING TEST
OF SPHERICAL TANK WITHOUT REINFORCEMENT
Tank A
MOVIE DURING SHAKING TEST
OF SPHERICAL TANK WITH REINFORCEMENT
Tank B
RESULT OF SEISMIC CAPACITY AND FRACTURE MODE
7000
Fracture
Responce Acceration αrmax (gal)
6000
reinforced
Fitting curve of Tank B
Trendline of Tank B
5000
Fracture
4000
Trendline of Tank A
Fitting curve of Tank A
3000
2000
Tank A JMA Kobe NS
Tank A Simulated seismic wave
Tank B JMA Kobe NS
Tank B Simulated seismic wave
Tank B sin wave
1000
0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
Maximum responce displacement Drmax (mm)
unreinforced
Broken intersection of braces
Broken brace around weld joint with upper column
reinforced
Broken brace around weld joint with lower column
SUMMARY（１／２）
The failure mechanism of the steel pipe brace structure of a spherical
tank were clarified by the elastic and elastic-plastic FEA.
From FEA results (with respect to the accident report), it was
concluded the failure mechanism of spherical tank as follows
(1) Seismic capacity of spherical tank with steel pipe brace, the
intersection of braces is the weakest point.
(2) The intersection of braces is in a multi-axial stress state due to the
structural characteristics of pipe brace configuration when seismic
inertia force is applied. This induces the fracture by plastic deformation.
(3) In case of long brace member is in tension, the intersection of long
brace member is received compression force from short brace
members, locally bending stress from cross-sectional deformation. In
consequence, the intersection of braces is deformed very large.
(4) The effect of the reinforcement by diaphragm and gusset plate were
discussed from FEA result of reinforcement models. by reinforcing the
intersection of braces, seismic resistance capacity were increased to
1.25 times by diaphragm, and to 1.39 times by gusset plate.
SUMMARY（２／２）
Non-linear response characteristics and failure mode of a spherical
tank were clarified by shaking tests using small models of spherical
tank.
From the test results, it was concluded the effectiveness of
reinforcement and the response characteristic of the pipe braced
supporting frame of a spherical tank as follows;
(1) In a case of brace intersection was reinforced, the maximum
response acceleration increased at collapse.
(2) In a case of there was reinforced there is no reinforcement to brace
intersection, structural strength indicated the lowest at the load
direction acting tensile force on long brace, cross section of long
brace were deformed largely.
ACKNOWLEDGMENT
This study used the results to verify the seismic
performance of the spherical tank in Research
Committee of the Ministry of Economy, Trade and
Industry commissioned project in 2014. Advice
and comments given by IHI Corporation has been
a great help in shaking tests. We would like to
express the deepest appreciation to committee
members and participator.
THANK YOU FOR YOUR KIND
ATTENTION.