Comparison between Japanese and North American method for

liquefaction assessment

Kanto Gakuin UniversityNozomu YOSHIDA

Introduction

What method do you use for liquefaction evaluation to average engineer? Japan: e.g. Design specification of Highway bridge U.S.: e.g. Technical paper by Seed 1971

►North American engineer studies more than Japanese engineer

►NSF Workshop in 1996 and 1998

Technical report NCEER-97-0022,Youd, T. L. etc., Journal of GT, Vol. 127, No. 10

Basic standpoint

North America United Engineers

►NSF Workshop Study and think

►Unsuggested (not recommended) issues

Japan Going my way

►Many design specifications Do not think or consider

►Everything is written◆ Do as written following the specification►Poor engineer education system

Highway bridge, Building foundation, Port facilities, Railway structures

Japanese academic system

Architectural Institute of Japan Architect Building engineer

Japan Society of Civil Engineers Road Airport and port Railway Dam .........

Governmental office Responsible only what they handles

►Do not like to follow outside organization

Why many specifications

When damage occurs, who is responsible? Japan

Engineer: I calculated following the design specification, therefore I am not responsible

Governmental office: I made it under the assistance of academic expert, therefore it is unexpected.

North America Sued by a customer?

►Moss Landing Marine Research Institute◆ damaged during 1989 eq.

Therefore, everything is to be written in the design specification. Otherwise somebody judged it resulting in responsibility

Compared specifications NSF workshop recommendation Highway bridge and Building foundation

Hwy. and Bulg.

Job or volunteer If job, revised on a periodic basis

Highway bridge by Public Work Research Institute If volunteer, may not revised without

something happen Building foundation

►1995 Kobe earthquake (Large ground shaking)►2011 Tohoku earthquake (Long duration)

Standard Penetration test

Turkey, Philippines Half of Japan

Recent auto or semi-auto machine Cone pulley: 63～73%

►With special care: 80～90% Semi automatic: 84% (average) Full automatic: 81% (Average)

ISO22476-3 Energy correction with measurement method

JIS A 1219（2013） No description, therefore no measurement method 8

Japan Hammer Falling Energy ratio

JapanDonut Free fall 78Donut Cone pulley 67

U.S.ASafety Cone pulley 60Donut Cone pulley 45

External load Bldg.: Equivalent cyclic

stress ratio

Hwy.: Stress ratio during an earthquake

CSR Cyclic stress ratio

0

0 0

av max vn d

v v

L r rg

τ α σσ σ

= =′ ′

0

0 0

0.65av max vd

v v

CSR rg

τ α σσ σ

= =′ ′

0

0 0

av max vd

v v

L rg

τ α σσ σ

= =′ ′

rn rn=0.1(M-1)

=0.65/MSF MSF: Moment scaling factor

►Number of effective cycles atτ av=0.65τmax

Reference liquefaction strength curve

Orion Bridge record during the 1971 San Fernando Eq.

0

0.2

0.4

0.6

0.8

1.0

10 5 3 1 0.5 0.3 0.1 0.050.03 0.01Equivalent number of cyclics at τ=0.65τmax

0 5 10 15 20 25Time (s)

-200

-100

0

100

200 Orion Bldg. NS, 1971 San Fernando Eq.

Bldg.

Positive NegativeAmp. N k Ne N k Ne

τmax 3.00 3.000.70τmax 2 1.20 2.40 2 1.20 2.400.65τmax 1 1.00 1.00 2 1.00 2.000.60τmax - - - 2 0.70 1.400.55τmax 6 0.40 2.40 - - -0.50τmax 1 0.20 0.20 3 0.20 0.600.45τmax 1 0.10 0.10 2 0.10 0.200.40τmax 2 0.04 0.08 3 0.04 0.120.35τmax 1 0.02 0.02 6 0.02 0.120.30τmax - - - - - -

Sum 6.20 9.84

N: Number of cyclesk: Conversion coef.Ne :Equivalent number of cycles

=N×k

Evaluation of effective number of cycles

Magnitude Ne at 0.65τmax

8.5 267.5 156.75 106 5~6

5.25 2~3

Acc.Max. dirPerp.All dataRep.

0

10

20

30

40

5 6 7 8 9Magnitude of earthquake

A-1

S-I

Mean

Mean + μ

Mean - μ

Line with gradient 0.2

can be approximated by

Magnitude Ne rn5.5 3 0.476.5 6 0.547 10 0.60

7.5 15 0.658.3 25 0.72

0.20.65( /15)n cr N=

0.1( 1)nr M= −

Applicability for M>8.5 Arai, 2011

Data scatters Has it meaning?

Moment magnitudeOther research shows 1/0.9～1/0.8Magnitude of earthquake

5 6 7 8 9 10

5

10

50

20

100

2011 Yume-no shima

2011 Chiba Port1987 Yume-no shima

1995 Kobe

Seed et al. (1983)AIJ (2001)

Bldg.

Hwy Shock and vibration type

►Number of cycles 2 and 3◆ Number of waves larger than τmax before τmax appears◆ Side same as τmax

rn equivalent value 0.55~0.70

MSF： Magnitude scaling factor （N.A.）7.5 7.5

/LCRR CRRF MSFCSR CSR MSF

= =

M Seed & Idriss Idriss*1 Ambra

seys

Arango Andrus &

Stokoe

Youd & Noble

Distance*2 Energy*3 PL<20% PL<32% PL<50%

5.5 1.43 2.20 2.86 3.00 2.20 2.8 2.86 3.42 4.446.0 1.32 1.76 2.20 2.00 1.65 2.1 1.93 2.35 2.926.5 1.19 1.44 1.69 1.60 1.40 1.6 1.34 1.66 1.997.0 1.08 1.10 1.30 1.25 1.10 1.25 1.00 1.20 1.397.5 1.00 1.00 1.00 1.00 1.00 1.00 - 1.008.0 0.94 0.84 0.67 0.75 0.85 0.87 - 0.73?8.5 0.89 0.72 0.44 - - 0.65? - 0.56?

*1 Lecture at 1995 Seed Memorial Lecture, no paper*2 Distance to liquefied site*3 Use energy from equivalent cycles by SeedToo conservativeRecommended (No recommendation for M<7.5)

Stress reduction coefficient, rd Apply until GL-20m Applicable until GL-

15m

Excel use

1 0.015dr z= − 1 0.00765 ( 9.15m)1.174 0.0267 (9.15m 23m)

d

d

r z zr z z

= − ≤= − < <

0.5 1.5

0.5 1.5 21 0.4133 0.04052 0.001735

1 0.4177 0.05729 0.006205 0.001210dz z zr

z z z z− + +=

− + − +

Predominant vibration modeDepth of engineering seismic bedrock

20

15

10

5

0

Dep

th (m

)

1.00.90.80.70.60.5rd

1-0.015z US-1 US-2rd=1-0.015z

0 0.5 1.00

10

20

30

Stress reduction coef. rd

Varorious ground

Average

2925

10

20

31

20

12

17

50+

27

6080

1020

10

50+

2048

1110

1250 9210

10 2020

1212

18

1230

20 30

302275

277560

2710

10

13 10

50+

31

16

8013

67

Fines content≥ 5%

Fc= 35% 15 ≤ 5

0 10 20 30 40 50

0.6

0.5

0.4

0.3

0.2

0.1

0

Pan America

Japan

China

No liq.Liq, Marginal

Chinese (Clay content =5%)

Corrected blow count (N1)60

τav

σ'v0

37

20

1510 26

4010

Recommendedin Workshop

SPT clean sand base curve

Liquefaction strength Bldg.

CRR Cyclic resistance

ratio

{ }1 60

7.5 21 60 1 60

( )1 50 134 ( ) 135 20010 ( ) 45

NCRRN N

= + + −− ⋅ +

16 16100

n

a ar

s

N NR aC

C

= +

Degree of liq. CommentSevere liq, Sand boil and ground subsidence more

than 2% or settlement of heavystructure more than 20cm

●

Medium liq. Sand boil and ground subsidence lessthan 2% or settlement of heavystructure less than 20cm

■

Border line Site to distinguish liq. and no liq. △

No liq. No sand boil nor subsidence ○

0 10 20 30 40 50Corrected N-value, Na

Shear strain amplitude

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

10% 5% 2%

Severe liq.Medium liq.Border lineNo liq.

No liq.

Liq.

Read datapoint fromfigure

Test vs Back Analysis

Basic concepts Relative density vs. strength

a=0.45，n=14 Value of C

Meyerhof

2 100

nd r r

o

D DaC

σσ

= + ′

97 19logaC DA= −94 19logsC γ= −

1170

70Nv

N C N Nσ

= =′ +

→ 116r fD N N= + Δ

* 1002170r

v

NDσ

=′ +

1 098/ vN Nσ′= ⋅

Confining stress dep.

Triaxial testSimple shear

0.0

0.5

1.0

DoubleAmp. (%) Sampling

FrozenConv.

DA=7.5%

DA=5%

DA=2.5%

2.5 5 7.5

Nc=15

σd Dr Dr n2σ'c 100 C= a{ +( ) }

0 20 80 10040 60Relative density, Dr (%)

γ=5%

Highway Bridge

After 1995 Kobe eq. liquefaction strength significantly changed based on frozen samples

0 5 10 15 20 25 30 0 5 10 15 20 25 300

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8Fine sandD50=0.15mmσ'v=1kgf/cm2

Fc=0%'

Fine sandD50=0.15mmσ'v=1kgf/cm2

Fc=30%'

Ⅳ

ⅠⅡⅢ

ⅠⅡⅢⅣ

Specification for highway bridges (1990 version)Specification for highway bridges (1996 version) (Cw=1)Design criteria of building foundation structures and commentariesDesign standard for railway structuresTechnical guidlines for aseismic design of nuclear power plantsTechnical standards for port and harbiur facilities in Japan

Liqu

efac

tion

stre

ngth

ratio

Liqu

efac

tion

stre

ngth

ratio

SPT N-value SPT N-value

Alluvial/Diluvial vs. Holocene/Pleistocene

In Japan, Alluvial and Diluvial are used instead of Geologic age (Holocene and Pleistocene)

In this presentation Holocene=Alluvial, Pleistocene= Diluvial

QuaternaryTertiaryCretaceousJurassicTriassicPermianCarboniferousDevonianSilurianOrdovicianCambrian

Precambrian

Cenozoicera

Mesozoicera

Palaeozoicera

0

65

245

540

10

20

0 50sea level (m)

Geological age

1,650

millenary

Sea level minimal,18 millenary years

100

2,580Change in 2010 byGoelocical Society of Japan

Hwy. 1996

Fill, ImprovedHolicenePleistocene

DA=5%Nc=20

Proposed cyclic stress rato by triaxial test, RL

0 0.2 0.4 0.6 0.8 1.00

0.2

0.4

0.6

0.8

1.0

0 10 20 30 400

0.2

0.4

0.6

0.8

1.0

Equivalent N-value, N1

Road Bridge

FillHolocenePliestocene

Frozen Tube

DA=5%Nc=20

Sandy soil

0 10 20 30 400

0.2

0.4

0.6

0.8

1.0

Equivalent N-value, N1

Fill, ImprovedHolocenePleistoceneShirasu

Sandy soilFc<10%

Fill DA=5%Nc=20

PleistoceneHolocene

Samping

1 1

6 4.51 1 1

0.0882 /1.7 ( 14)

0.0882 /1.7 1.57 10 ( 14) ( 14)L

N NR

N N N−

≤= + ⋅ ⋅ − >

1 1 2

1

1

2

170 /( 70)1 (0% 10%)

( 40) / 50 (10% 60%)/ 20 1 (60% )

0 (0% 10%)( 10) /18 (10% )

a

v

c

c c

c c

c

c c

N C N CN N

FC F F

F F

FC

F F

σ= +

′= +≤ <

= + ≤ < − ≤

≤ <= − ≤

Hwy. 2017

6 4.5

0.0882 (0.85 2.1) /1.7 ( 14)( 14)0.0882 /1.7 1.6 10 ( 14)

a aL

aa a

N NR

NN N−

+ <= ≥+ × ⋅ −

1 50

10 50 1 50

1

( 2.47) 2.47 ( 2mm){1 0.36log ( / 2)} ( 2mm)

170 /( 70)1 (0% 10%)

( 20) / 30 (10% 40%)( 16) /12 (40% )

FCa

v

c

FC c c

c c

C N DN

D N DN N

Fc F F

F F

σ

+ − <= − ⋅ ≥

′= +≤ <

= + ≤ < − ≤

0 10 20 30 40

0.2

0.4

0.6

0.8

1.0

Equivalent N-value, N1

Road Bridge

FillHolocenePliestocene

Frozen Tube

DA=5%Nc=20

Sandy soil

Samping

2017-20

0

20

100500Fc (%)

N1=15

0.1

-2.47

Findings on frozen sampling

Frozen sample has been believed to be an undisturbed sample

10 100 10000

0.5

1.0

1.5

2.0

2.5Holocene sandPleistocene sandHolocene clayPleistocene clayFrozen sample

In-situ shear modulus, GOF (MN/m2)

10 100 1000

原位置せん断弾性定数 , GOF (MN/m2)

0

0.5

1.0

1.5

HolocenePliestocene

Frozen TubeSamping

Past researches New tests, PWRI

Liquefaction strength

Corrected N value from free fall method (N1)78

Corrected N value from pulley method, (N1)65

Liqu

efac

tion

stre

ngth

σd /

(2σ'

0)

Froz

en sa

mpl

e sp

ecim

en

0 10 20 30 40 50 60

10 20 30 40 5000

0.2

0.4

0.6

0.8

1.0

Tube sample specimen

Water pressurepistonRotally typedouble tube

0 10 20 30 400

0.2

0.4

0.6

0.8

1.0

Equivalent N-value, N1

Road Bridge

FillHolocenePliestocene

Frozen Tube

DA=5%Nc=20

Sandy soil

Samping

Past researches New tests, PWRI

Fines contents

Bldg.

Fc (%) ΔNf0～5 05～10 Interpolation10～ 0.1Fc

←OriginalRevised based onrecent research 0 10 20 30 40 50

0

2

4

6

8

10

12

Fines content (%)

Equivalent N-value, N1

0

0.2

0.3

0.4

0.6

0.7

0.5

0.1

0 10 20 40 5030

γ=2%5%10%

DA=5% (γ=3.8%)Frozen

DA=5% (γ=3.8%)Existing

Fc≤ 5%

Iwasaki et al.

Equivalent N-value, N1

0

0.2

0.3

0.4

0.6

0.7

0.5

0.1

0 10 20 40 5030

γ=2%5%10%

Fc≥ 10%

Severe liq.Medium liq.Border lineNo liq.

Parallel

Highway Bridge

0 20 40 60 80 100-0.4

-0.2

0

0.2

0.4

細粒分含有率 Fc（%）

有効拘束圧 σ'c (kPa)100<σ'c≤20050<σ'c≤10020<σ'c≤50

σ'c≤200

0 20 40 60 80 1000

5

10

細粒分含有率 Fc (%)

1( 2.47) 2.47a FCN C N= + −

Nc=20

平均

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.01 0.02 0.05 0.1 0.2 0.5 1 2 5D50 (mm)

Sites DA (%) σ'c (kPa)A 6 20~115B 5 45~175C 5 150D 5 50E 6 37~77F 6 50, 100G 6 80~120

1980version

1980 1990

2017200

100

02520151050

Fc=100%90%80%70%60%

Fc= 50%40%30%20%0～10%

N1

North America1 60( )aN Nα β= +

Parameter Fc≤5% 5%<Fc<35% 35%≤Fc

α 0 5

β 1 1.2

21.76 190/ cFe −

1.50.99 /1000cF+

Various factors (N.A.)Factor Variable Nt. Correction

Overburden stress － cN (Pa/σ'v0)0.5

Overburden stress － cN CN ≤1.7Energy ratio Donut cE 0.5～1.0Energy ratio Safety cE 0.7～1.2

Energy ratio Automatic fall donuts cE 0.8～1.3

Diameter of borehole 65～115mm cB 1.0

Diameter of borehole 150mm cB 1.05

Diameter of borehole 200mm cB 1.15

Rod length <3m cR 0.75Rod length 3～4m cR 0.8Rod length 4～6m cR 0.85Rod length 6～10m cR 0.95Rod length 10～30m cR 1.0Sampling method Standard cs 1.0Sampling method No liner cs 1.1～1.3

1 60( ) N E B R sN c c c c c N=

0

2.21.2 /N

v a

cPσ

=′+

CN: 2→1.7Better eq.

Pa=101 kPa

Japan

98N

v

cσ

=′

17070N

v

Cσ

=′ +

Other (N.A.)

Seed (Original)

Kσ: Confining stress correction Recommended among

various research

7.5L

CRRF MSF K KCSR σ α=

10( / ) f

v aK Pσ σ −′=Dr≤40% (f=0.8)

Kσ=(σ'v0/Pa)f-1

0 1 2 3 4 5 6 7 8 9 10Effective overburden stress ratio σ'v0/Pa

0

0.2

0.4

0.6

0.8

1.0

1.2

Kσ

Dr≥80% (f=0.6)

Dr≈60% (f=0.7)

Correction by slope Defined ad

Determine by triaxial test, but large scatter►Average engineer do not use

Aging effect Seed: 25% increase in 100 days Youd: Young ground is more liquefiable Not recommended because of short data Old sediment (older than several thousands)

►Limited engineer uses aging, not Kσ

0/st vα τ σ′=

Design ground shaking N.A.

Consider only Magnitude and other factor such as area, duration, fault mechanism is difficult. Conservative side

►Not a big issue in the liquefied site Use Moment magnitude, Mw

PGA when liquefaction does not occur►Empirical equation considering earthquake magnitude,

focal distance, site condition, etc.►If empirical eq. is not available, seismic response analysis

such as SHAKE and DESRA►Use amplification factor to be multiplied to PGA at the

engineering seismic base layer◆ Require highly engineering judgement

2 directional components►geometric mean, but larger value is conservative

High frequency component (Period<0.1 s）►Spiky wave does not cause displacement because of short

active time, therefore neglect►High frequency component is attenuated in SHAKE and

DESRA►When using amplification factor, choice of frequency

range is important

Japan Bldg. (1985）

►Affected by various factors◆ Some of them is not clear◆ Affected by local ground condition If liquefied, earthquake motion does not propagate to

the ground surface►Target is horizontally layered deposit, but important is

the case with structure exists►Proposed method gives rough indication►THEN, PGA recorded during past earthquakes is relevant◆ 200 cm/s2

Kawagishi-cho apartment house in Niigata eq. = 158

Bldg. （2001）►PGA is a result of response of ground, and is affected by

the ground conditions.►Damage limit: 150~200 cm/s2

►Ultimate limit: 350cm/s2◆ PGA at Port Island during the 1995 Kobe eq.

PGA in liquefied site

Hwy 1996►Change of acceleration by liquefaction is not considered

in αmax

►Considering liquefaction requires effective stress seismic response analysis, but it is impossible

►FL method is a simplified method►FL method is safety factor method◆ External load becomes large under larger ground shaking

PGA in non liquefied case~1995 1996~

0.15

Ground type 1 2 3

Type 1 (Ocean trench) 0.3 0.35 0.4

Type 2 (Near field eq.) 0.8 0.7 0.6

Seismic coefficient

Only level 2 eq.

Comparison of Japanese specifications

35

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

砂質土粘性土シルト礫腐植土その他

0 50N 値

Comparison between Bldg. and Hwy

1996 version

0.1

1

10

0.1 1 10FL,ROAD

0≤Fc≤1010<Fc≤2020<Fc≤4040<Fc

FL,ROAD

N≤1010<N≤20N>20

0.1

1

10

0.1 1 10

2017 version

0.1

1

10

0.1 1 10FL,ROAD

0≤Fc≤1010<Fc≤2020<Fc≤4040<Fc

FL,ROAD

N≤1010<N≤20N>20

0.1

1

10

0.1 1 10

-20

0

20

100500Fc (%)

N1=15

Definition of FL

Bldg.

Hwy vs. Budg

38

0

0

max vn d

v

L r rg

α σσ

=′

0 .1( 1)nr M= −

1 2 3 4 5 1 52 1 5

2/rL L L

Ln

c c c c c R c c R R RFL L

c c CcrL Lc

= = = =

01 20 .9 0 .573r

KC += =

Relative density, Dr

SPT N Dr（%）φ (deg.)

Peck Meyerhof

0～4 very loose 0～20 <28.5 <30

4～10 loose 20～40 28.5～30 30～35

10～30 medium 40～60 30～36 35～40

30～50 dense 60～80 36～41 40～45

50以上 very dense 80～100 >41 >45

1002170r

v

NDσ

=′ +

Meyerhof

Relative density in sites

150

100

50

0

N値86420

0.4

0.3

0.2

0.1

0.0

平均相対密度 Dr (%)100500

粒分含有率 Fc（%）

相対密度

Dr（

%）

0 10 30 40 5020

20

80

0

40

60

100

120

140

相対密度

Dr (

%)

N値

120

100

80

60

40

20

0706050403020100

Miyagi Pref.

Seed

MikamiRL vs. Dr

Accuracy (2011, PWRI)

Liq. No Lq.FL≤1 53 35FL>1 0 24

Tokyo Bay area: Old fill / natural deposit Many no liq. site although FL≤1 Fill after 1945 liquefy There is no clear difference between borehole data

between liq. and no liq. sites Tone River area

FL is relatively large in the liquefied sites Thin thickness in case FL≤1 41

Accuracy (2007)

-400

-200

0

200

400

20151050

GL-32m, NS

Time (s)

Port Island

600550500450400350300250200150100500-200

-100

0

100

200

Time (s)

Tonankai

-100

-50

0

50

100

200150100500Time (s)

Tomakomai

Average Std. dev.

Port Island

Ground number

2.0

1.5

1.0

0.5

0.0250200150100500

Average Std. dev.

Tonankai

Ground number250200150100500

2.0

1.5

1.0

0.5

0.0

Average Std. dev.

2.0

1.5

1.0

0.5

0.0

Tomakomai

250200150100500Ground number

FL value

1.0

0.5

0.0

FL value

1.0

0.5

0.0

Port Island

543210FL value

1.0

0.5

0.0

Tonankai

86420

Tomakomai

86420

YUSAYUSAL N

LN

YUSAYUSAL N

LN

YUSAYUSAL N

LN

Proposed correction, c2=0.5 is too conservative! Large scatter

under ocean trench type eq.

PI Tonankai TomakomaiLiq. both 64.7 6.9 4.2

Conservative 15.0 0.4 4.9hazard ratio 1.4 46.8 11.6no liq. both 18.9 45.9 79.3

Accuracy rate 83.6 52.8 83.5

TNK TMK51.7 15.834.1 73.82.1 0.112.2 10.363.9 26.1

Depth

(m)

SoilType

Vs

(m/s)

g t

(k N/m3)

Max. Excess PWP

100 200

Max. Stress(kPa)

20 40 60

(kPa)

19.1

19.1

19.119.1

20.1

20.6

19.1

17.1

17.118.1

18.1

18.1

18.117.119.1

190.5

190.5

190.5145.4

141.2

225.8

166.4

127.0

127.0217.2

217.2

172.4

220.7217.2294.7

0.8

2.43.03.8

5.5

7.8

9.2

12.813.2

15.0

16.8

18.118.919.520.5

Depth

(m)

SoilTy pe

Vs

(m/s)

gt

(kN/m3)

Max . Stress(kPa)

20 40 60 80

Max. Excess PWP

100

PortIsland東南海苫小牧

100181.6

181.6

181.6

271.3

271.3

271.3

271.3

19.6

19.6

19.6

18.6

18.6

18.6

18.6

0.5

2.0

9.9

12.1

13.7

15.5

17.9

(kPa)50 150

rd is good evaluation

Concluding remarks

Same framework, but different definition Liquefaction strength

Average or boundary►If average, half of them is in critical side!◆ Result was conservative, Why?

Parameters SPT N-value, overburden stress, fines contents

►Is those sufficient? No!◆ What are other parameters? Aging, Kα, ....