Trend of design codes of the geotechnical engineering field in Japan

1
Trend of design codes of the geotechnical
engineering field in Japan

• Makoto Suzuki
• Shimizu Corporation

2
Two topics

• Introduction of design codes toward
  performance-based design in Japan
• Harmonization among Design Codes in the Asian
  Region (ACECC Workshop)
• Application on partial factor design
• Code Calibration of Designing Pile Foundations
  based on the Partial Factor Approach

3
Harmonization among Design Codes within the Asian
Region in geotechnical field
4
Contents

• Background of design codes
• Toward performance-based design
• Eurocode 7
• Japanese activities
• JGS Standard

5
Background of Design Codes

• WTO/TBT Agreement in the 1990s recognized the
  necessity of standardization of quality design.
• It was becoming increasingly important to respect
  the international standards required by the
  WTO/TBT Agreement and formulate performance
  specifications.
• Japan should adopt the limit state design methods
  indicated in ISO2394
• Seismic design efforts to disseminate information
  in the international community would encourage
  the progress of geotechnical engineering in
  Japan.

6
Contradictory International Trends

• ISO and Eurocodes in Europe
• international and regional standardization,
  unification
• Seismic performance design in North America
• to enhance the degree of freedom in design

7
European Nations

• ISOISO2394
• General principles on reliability for structures
• CENEN1990
• Eurocode - Basis of structural design
• Vienna Agreements
• WTO/TBT Agreement

8
Eurocode 7

• Eurocode 7 (EN1997) is a standard for soils and
  foundations.
• stabilization (ultimate limit state ULS) and
  elasticity (serviceability limit state SLS) by
  Terzaghi
• Terzaghi and Peck utilized the safety factor in
  design codes.
• Brinch Hansen brought in the partial factor
  design.
• CEN/SC7 created the first partial factor method
  which was unique to geotechnical engineering.

9
Eurocode 7

• While Eurocodes are partial factor methods
  expected to use the safety margin as load
  factors, EC7 focus on material factors.
• Large influence of the uncertainty of the soil
  has to be considered.
• EC7 could not conform with other materials.
• Two separate calculations of structures and soils
  using different partial factors became necessary
  in foundation designing.
• Three approaches have been adopted.

10
Japanese Activities in JGS

• 1997-1999
• Committee to Study the Current Status Foundation
  Design in Japan Its Future, chaired by Prof.
  Kusakabe
• Comprehensive foundation design code Geocode
  21
• 2000
• Foundation Design Standard Committee, chaired by
  Prof. Kusakabe
• Draft JGS standard were prepared
• 2001-2003
• Committee on Standardization of Foundation
  Design, chaired by Prof. Honjo
• English translation of Geocode 21 ver.2.0

11
Japanese Activities in JGS

• 2004
• Committee on Geotechnical Design and Construction
  Standards
• Principles for Foundation Design Grounded on
  Performance-based Design Concept (tentative)
• JGS Standard JGS4001-2004

12
Japanese Activities

• 2002 MLIT
• Committee on the Basis of Structural Design for
  Buildings and Public Works at the Japan Institute
  of Construction Engineering
• "Basis of Structural Design for Buildings and
  Public Works, published by MLIT(the Ministry of
  Land, Infrastructure and Transport)
• 2001-2003 JSCE
• Basic Research Committee on Comprehensive Design
  Code Development, chaired by Prof. Kusakabe
• Principles, Guidelines and Terminologies for
  Structural Design Code Drafting Founded on the
  Performance-based Design Concept ver.1.0 (code
  PLATFORM)

13
JGS Standard
Principles for Foundation Design Grounded on
Performance-based Design Concept

• A4 size 246 pages \2,835
• member \1,890
• Contents
• -standard text (chapter 0-7)
• -commentary
• -appendix
• -standard text in English

14
Chapter setting up in present and the future

• BASIS OF STRUCTURAL DESIGN
• BASIS OF DESIGN OF FOUNDATION STRUCTURES
• GEOTECHNICAL INFORMATION
• DESIGN OF SHALLOW FOUNDATIONS
• DESIGN OF PILE FOUNDATIONS
• DESIGN OF COLUMN-TYPE FOUNDATIONS
• DESIGN OF RETAINING STRUCTURES
• DESIGNING TEMPORARY STRUCTURES
• EMBANKMENT
• SLOPE (or CUTTING)
• TUNNEL, CAVARN
• GROUND IMPROVEMENT
• REINFORCED SOIL

chapter 0 - chapter 7 in present
on going
future
15
Composition of the standard

• Performance requirements were hierarchically
  organized.
• The hierarchy of descriptions is three-tiered
  purpose, required performance, and performance
  specifications.
• Two approaches approach A and approach B

16
Problem Items

• Standardization of characteristic values of
  geotechnical parameters
• They were defined as the averages of derived
  values.
• Performance regulations and verification
• The method of the performance regulations
  according to the verification technique.
• What is the verification approach A?
• Checking list for each foundation designing
• From foundations to earth structures
• No load specification
• Qualifications of engineers

17
Conclusion

• JGS Standard is a set of design principles
  concerning foundation structures.
• There is a subcommittee on the performance
  evaluation of earth structure in JSCE.
• ISSMGE's ITC23 (Limit State Design in
  Geotechnical Engineering Practice, chaired by
  Prof. Honjo)

18
Code Calibration of Designing Pile Foundations
based on the Partial Factor Approach
19
Background

• Cordification of structural design
• Performance and world-wide standardization
• Partial factor design format
• Revision for compliance with ISO standard

20
Contents

• Partial factor design for infrastructures
• Basic variables of design verification format
  considering partial factors
• Evaluation of the reliability indexes of existing
  structures in terms of safety margin
• Determination of partial factors from the result
• Verification of the effectiveness of partial
  factor design

21
LSD Reliability-based design

• Structural design is performed under
  uncertainties.
• Probabilistic description of safety margin is
  rational.
• Safety factor is not quantitative index.

Nominal deterministic load effect
Deterministic resistance
PDF
S
R
Load effect (unknown)
Pf
Load effect / Resistance
22
Levels of Reliability-based Design

• Level I
• ? Partial factor design
• Level II
• ? Design based on reliability index b
• Level III
• ? Design based on probability of failure Pf

easy
hard
23
Reliability analysis Probability of failure
(Level 3)
Performance function of bi-variates
Load effect fS(x)
f
x
f
x
(
),
(
)
R
S
Resistance fR(x)
If there are normal distributions
r
s
,
Performance function of multi-variates
Difficult in general
Volume integral of fX(x)
24
Approximate Method of Analysis

• Level III is generally difficult.
• Is a statistical character of the random variable
  certain?
• Is it possible to calculate from a complex
  performance function?
• Introduction of linear approximation (Level II)
• only the second moment of the random variable
• linear approximation (normal dist.)
• most probable failure point

25
FORM Reliability index
(Level 2)
f
(
m
)
bs
M
Reliability index
FORM
M
or
failure
safety
FOSM
M
M
lt0
gt0
Safety margin M
s
s
M
M
P
f
m
m
0
M
Approximate technique
General linearization
26
Reliability Index in Normalized Space
z
2
x
2
normalized space
z
b
1
m
z
z
(
,
)
2
1
2
x
x
(
,
)
1
2
original space
m
x
1
1
most probable failure point
27
Sensitivities of Random Variables
sensitivities
a
i
Sensitivities utilize to set partial factors
28
Normal Distribution Approximation
Rackwitz Fiessler
1) same probability 2) same density
f
x
(
)
X
j
j
Equivalent normal distribution
Arbitrary probability distribution
x
m
x
j
j
from reliability index to probability of
failure
FORM
29
What is a partial factor design?

• Reliability-based design at level 1
• Probabilistic limit state design
• Not performance function, but design verification
  function

material property of element i partial factor of
element i represent value of load j partial
factor of load j
30
Code calibration

• Design value method
• Reliability Sensitivities according to FORM
• Relatively easy way
• Experiences for code calibration

Design value
characteristic value
design value
Recommennded standard sensitivities in Eurocodes
partial factor
Dominating resistance Other resistance Dominating
load Other load
g
R
/ R

m
k
d
g
S
/ S
f
d
k
31
Study flow

• For the reliability assessment of existing
  structures, values of reliability indices are
  calculated by using FORM.
• Sensitivities of input parameters can be obtained
  at the same time.
• Partial factor format is studied.
• Target reliability index is determined and
  partial factors are calculated using
  sensitivities and target reliability index.
• Verification of the effectiveness of this study
  is performed comparing between reliability
  indices of case design and target reliability
  indices.

32
Open-type wharf on vertical piles
Twenty existing structures are selected for the
calibration
Steel pipe pile
33
Action on wharves
Dead load Surcharge
Seismic events
During an earthquake - Dead weight of
superstructure 20 kN/m2 -
Surcharge 10 kN/m2 - Lateral
seismic coefficients 0.05
0.25 During berthing - Dead weight of
superstructure 20 kN/m2 -
Surcharge 20 kN/m2 - Reaction
force of fender depends on the design
water depth
Inertia force at earthquake
Berthing events
Dead load Surcharge
Reaction force of fender
34
Analytical method

• A push-over analysis, in which lateral load is
  gradually increased in a performance function, is
  conducted.
• A bilinear model, in terms of the relationship
  between the bending moment and curvature of the
  pile, is used.
• When a lateral load subjects, yielding generally
  occurs in the pile top at first. Then, yielding
  occurs in the underground portion of the pile,
  and the ultimate bearing capacity of the pile is
  exceeded.
• By using the advanced first-order second-moment
  method, the safety of the structure was evaluated
  in terms of reliability index ?.

35
Limit states performance function

• Serviceability limit state on flexural strength
  of pile
• Repairability limit state on flexural strength of
  pile
• Ultimate limit state on bearing capacity of pile
• fy yield strength of steel pile
• kh coefficient of lateral subgrade
  reaction
• Rb, Rf bearing capacity of pile
  tip and skin friction of pile

36
Statistics of basic variables
the coefficient of subgrade reaction
kh 2000N (kN/m3) COV of the coefficient of
subgrade reaction kh 0.758 - 0.817
37
Reliability index b for existing wharves

• (flexural strength of pile)

8.53
6.49
bT6.0
bT3.0
38
Reliability index b for existing wharves

• (bearing capacity of pile)

bT3.0
39
Sensitivity of input parameters
(SLS RLS on the flexural strength of pile)
fy
0.95
0.87
0.18
0.34
Yield strength of pile (SLS)
kh
Coefficient of horizontal subgrade reaction(SLS)
(a) Berthing (SLS)
(b) Seismic event (SLS)
fy
0.96
0.08
kh
afy is prominent
akh is notable
(c) Seismic event (RLS)
40
Sensitivity of input parameters
(ULS on the bearing capacity of pile)
Rb
Bearing capacity
0.83
Bearing capacity
0.53
Rf
Yield strength of pile
Yield strength of pile
Skin friction
Skin friction
Coefficient of horizontal subgrade reaction
Coefficient of horizontal subgrade reaction
(e) Seismic event
(d) Berthing
The effects of fy kh are minimal in either
case. The sensitivities of Rb Rf vary with
pile diameter length.
41
Relationship between aRb and aRf
aRb0.83 aRf0.53
? berthing ? seismic
42
Condition of setting partial factor

• Normal distribution
• Log-normal distribution
• Uniform set of sensitivities

Appropriate safety margins
43
Reliability index values obtained using proposed
partial factors
(flexural strength of pile)
(a) Berthing SLS of flexural strength of pile
(b) Seismic SLS of flexural strength of pile
Some cases are below the target b
(c) Seismic RLS of flexural strength of pile
44
Reliability index values obtained using proposed
partial factors
(Bearing capacity of pile)
(d) Berthing ULS of bearing capacity of pile
(e) Seismic ULS of bearing capacity of pile
45
Conclusions

• For wharves, three limit states were defined for
  berthing and seismic event. The partial factor
  method was proposed for the flexural strength and
  bearing capacity of the piles.
• The reliability of 20 existing wharves designed
  according to the current code was evaluated, and
  partial factors were determined based on the
  sensitivities of basic variables.
• The structure was redesigned using the partial
  factor method to verify the effectiveness.
• The reliability evaluation is conditional,
  because design loads based on the current code
  were used. When reviewing load factors, a
  discussion from the viewpoint of the probability
  of occurrence is necessary.