Sei sulla pagina 1di 8

INDIAN GEOTECHNICAL SOCIETY, KOLKATA CHAPTER

GEOTECHNICS FOR INFRASTRUCTURE DEVELOPMENT

KOLKATA 11th – 12th March 2016, Kolkata, West Bengal, India

GROUND RESPONSE AND LIQUEFACTION ANALYSIS


Tanumaya Mitra, Post Graduate Student, IIEST Shibpur, tanumayamitra@gmail.com
Ambarish Ghosh, Professor of Civil Engineering, IIEST Shibpur, ambarish@civil.iiests.ac.in

ABSTRACT: Kolkata, one of the oldest metropolitan cities of India, located on the eastern bank of river Hooghly,
lying on seismic zone III as per the seismic zonation map of IS 1893:2002 has suffered a number of damages
inflicted by several earthquakes in the past. In the present study, two borehole data have been collected to analyze
seismic liquefaction hazard of the region. Input motion for ground response analysis has been obtained from
SEISMOMATCH software. The ground motion parameters viz. PGA and SA at the ground surface have been
estimated by using 1D seismic ground response analysis software DEEPSOIL. The PGAs obtained from the
DEEPSOIL software for different locations have been used to determine the probable depth of liquefaction and
Liquefaction Potential Index (LPI) as per as the latest state of the art methodology for Mw 6.7. Liquefaction
analysis has been carried out using MATLAB.

INTRODUCTION
Soil Liquefaction has been one of the major SELECTION OF GROUND MOTION
problems in Geotechnical Earthquake Engineering. The peak ground acceleration can be estimated by
Its devastating effects sprang to the attention of using Ground Motion Predictive Equations
geotechnical engineers in a three-month period in (GMPE). However no such reliable attenuation
1964 when the Good Friday earthquake (MW=9.2) relationship is available for Kolkata. According to
in Alaska was followed by the Niigata earthquake seismic hazard map of India, the city is located in
(MS=7.5) in Japan. Both earthquakes produced moderate seismic zone (Zone III) with a zone
spectacular examples of liquefaction-induced factor of 0.16. In this area, no seismic station is
damage, including slope failures, bridge and present and no records of strong ground motion
building foundation failures and flotation of buried are available. 1D ground response analysis has
structures. Hence, evaluation of soil liquefaction been carried out in DEEPSOIL software using two
resistance is an important aspect of geotechnical different Peak Bedrock Acceleration values of 0.10
engineering practice. Liquefaction analysis g and 0.16 g using Equivalent Linear analysis.
framework was first proposed by Seed and Idriss Input motion with PBRA value of 0.1 g has been
(1971) and it is well known as the ‘Simplified generated by direct scaling of Imperial Valley
Procedure’ which determines the factors of safety Earthquake motion and Input motion with PBRA
against liquefaction by taking the ratio of capacity value of 0.16 g has been generated by spectral
of a soil element to resist liquefaction to the matching.
seismic demand imposed on it. Capacity to resist
liquefaction is computed as the cyclic resistance TARGET BEDROCK SPECTRUM AND
ratio (CRR) and seismic demand is computed as SPECTRAL MATCHING
the cyclic stress ratio (CSR). Factor of safety for a The target bedrock spectrum has been obtained
soil layer can be calculated with the several in-situ from IS 1893:2002 based on the desired PBRA at
tests such as SPT, CPT, BPT and shear wave 5% damping for rock or hard soil. Spectral
velocity test (Vs) test (Youd et al.,2001) . In this Matching is based on frequency domain techniques
study SPT based liquefaction analysis framework that alter the Fourier amplitude (but not phase)
has been used to determine the variation of Factor spectrum of the original ground motion so as to
of Safety with depth. match a target response spectrum. This approach
Tanumaya Mitra, Ambarish Ghosh

suffers because it changes the non-stationary nature

Acceleration (m/s2)
Design
of the records. Lilhanand and Tseng (1988) Accelerogram
introduced a perturbation approach based on
Duhamel’s integral that modifies the original signal Original
into spectrum matching one. The perturbation Accelerogram

added was localised in time so that it attempted to


maintain similar non-stationary statistics to the
original. Abrahamson(1992) developed the Time (sec)
approach discussed in Lilhanand and Tseng (1988) Fig.4 Comparison of Original and Generated
and produced Rspmatch software which was Accelerogram
further developed by Hancock et. al.,(2006) . This
SCALING OF INPUT MOTION
algorithm is embedded in Seismomatch software.
Input motion with PBRA value of 0.1 g has been
generated by direct scaling of Imperial Valley
Earthquake motion.
Acceleration (m/s2)

0.2
Imperi

Acceleration (m/s2)
A c c e le r a t i o n ( g )

0.1
0
-0.1
-0.2
-0.3
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38
Time
Time (sec) Time (sec)

Fig.1 Original Accelerogram of Imperial Valley Fig.5 Accelerogram of Imperial Valley


Earthquake Earthquake Scaled down to 0.1 g
1.5
1.4
1.3 Target Spectrum

1.2
1.1
Imperi
GROUND RESPONSE ANALYSIS
1
Influence of near surface geological conditions in
Ac celer ation (g)

0.9
Sa(g)

0.8
0.7 the form of sediment amplification or site response
0.6

0.5
0.4
is apparent from the damage distribution of many
0.3
0.2
destructive earthquakes (Edward H.,1996). It is a
0.1
0
0 1 2 3 4
proven fact that earthquake ground motion alters as
Period (sec)

Period (sec) it propagates through the soil from bedrock to the


surface. The soil acts as a filter and changes the
Fig.2 Target Spectra and Response Spectra of frequency content of the motion. Earthquake
original accelerogram damage is a function of the magnitude or the
energy released, which in turn depends on the
0.45
Target Spectrum
source, path and site. Magnitude of earthquake,
0.4

0.35
Imperi
degree of shaking, and destruction caused is
0.3 dependent on several factors. Magnitude of
Ac c eleration (g)
Sa(g)

0.25

0.2
earthquake is proportional to the energy released
0.15 and it might attenuate or amplify as it travels away
0.1

0.05
and spread over larger province. The degree of
0
0 0.5 1 1.5 2
Period (sec)
2.5 3 3.5 4
shaking of ground relies on the matching of the
Period (sec) fundamental frequency of ground and the building
and the degree of damage of structure is in turn
Fig.3 Spectrally matched ordinates of influenced by the properties and type of rock, soil
generated accelerogram deposits, tectonic and geomorphological features.
Susceptibility of subsoil to liquefaction, a complex
INDIAN GEOTECHNICAL SOCIETY, KOLKATA CHAPTER
GEOTECHNICS FOR INFRASTRUCTURE DEVELOPMENT

KOLKATA 11th – 12th March 2016, Kolkata, West Bengal, India

behavior of soil due to the decrease in effective


stress, is also regulated by degree of shaking. Table 1 Soil profile – BH1 Dumdum
Hence, it is essential to carry out ground response
analysis to determine the ground motion Layer d
γ
Vs PI
Soil Type (kN/ N
parameters. Ground response analysis techniques No. (m) (m/s) (%)
m3)
avg

are often categorized based on the dimensionality


of the problem and the characteristics of the soil 1 4.5
Soft Silty
16.5 3 118 10
model used in the analysis (Kramer 1996). The clay
problem can be 1D, 2D and 3D, and can be solved Soft Silty
2 7.5 16.5 2 110 15
by linear, equivalent linear, and nonlinear clay
approaches. The linear approach requires constant Stiff silty
3 7.5 18 8 155 38
values of shear modulus (G) and damping ratio (ξ) clay
for an induced level of small shear strain in each 4 12 Silty sand 22 52 400 0
layer, while the nonlinear approach uses cyclic Hard Silty
5 7.5 22 27 360 20
stress-strain models (backbone curve) in order to clay
capture the irregularity in the material property. In 6 6 Silty sand 22 58 650 0
the equivalent-linear approach the nonlinearity of
the shear modulus and damping with cyclic shear
Table 2 Soil profile – BH2 Tollygunge
strain is considered by using the equivalent-linear
soil properties by an iterative procedure, to obtain
values of the shear modulus and damping ratio, γ
Layer N Vs PI
which are compatible with the effective shear d(m) Soil Type (kN/
No. (m/s) (%)
m3)
avg

strains in each layer.


Soft Silty
1 6 16.5 2 110 5
DYNAMIC SOIL PROPERTIES clay
A set of material curves have been defined in 2 7.5
Soft Silty
14 2 120 5
DEEPSOIL for defining strain dependent shear clay
modulus (modulus reduction curves) and damping Stiff silty
3 7.5 17 11 220 10
ratio for different soils. In the absence of site clay
specific modulus reduction and damping ratio 4 7.5 Silty sand 20 36 325 0
curves, standard curves proposed by Vucetic and
Dobry (for clay) and Seed and Idriss (for sand) that Stiff Silty
5 9 18.5 22 385 24
are being extensively used in several studies, are clay
considered to be the better alternative. Shear wave Hard silty
6 6 19.5 47 630 24
clay
velocity values measured in the field have been
used for the ground response analysis. Water table Dense
7 11.5 20 80 720 0
has been considered at the ground surface. silty sand

SOIL PROFILES USED FOR STUDY RESPONSE SPECTRA


Two borehole data have been collected from Response spectra at the ground surface have been
Dumdum and Tollygunge area. Soil profile for obtained for two different PBRA values of 0.10 g
both Dumdum and Tollygunge consist of silty and 0.16 g. It has been observed that the PGA
clay/clayey silt upto a depth of about 20 m and value gets amplified while passing through the soft
then a layer of medium dense to dense silty sand soil deposit. Table 3 shows the summary of ground
upto about 30 m. response analysis.
Tanumaya Mitra, Ambarish Ghosh

LIQUEFACTION ANALYSIS
Table 3 Summary of Ground Response
Analysis
Determination of Cyclic Stress Ratio
Cyclic Stress Ratio (CSR) characterizes the seismic
Site PBRA PGA PBRA PGA
demand induced by a given eartquake and can be
Dumdum 0.178 g 0.255 g determined from peak ground acceleration that
0.10 g 0.16 g depends on site specific ground motions. The
expression for CSR is given as:
Tollygunge 0.151 g 0.205 g

CSR = 0.65 rd ………………….(1)

1.2
PBRA 0.10 g 0.65 is a reference stress level (Seed and Idriss
1 PBRA 0.16 g 1971); amax is the peak horizontal ground
acceleration; g is the acceleration due to gravity; σv
Spectral Acceleration (g)

0.8 and σv ‘ are the total vertical overburden stress


and effective vertical overburden pressure at a
0.6
given depth below the ground surface; rd is the
0.4 stress reduction factor that accounts for the
dynamic response of the soil profile; MSF is
0.2 magnitude scaling factor and kσ is the
0
overburden stress correction factor.
The stress reduction factor, rd, is used to determine
0 5 10 15
the maximum shear stress at different depths in the
Period (sec) soil. Values generally range from higher at the
ground surface to lower values at larger depths.
Fig.6 Response spectra at the ground surface Idriss (1999) performed several hundred
for BH 1 Dumdum parametric site response analyses and concluded
that, for the purpose of developing liquefaction
evaluation procedures, the parameter rd could be
1
PBRA 0.10 g expressed as:
0.9
PBRA 0.16 g
0.8
Spectral Acceleration (g)

rd = exp [α(z) + β(z) × M] …………………(2a)


0.7
&
0.6
α(z) = −1.012 − 1.126 × sin ( * 5.133. …(2b)
0.5 ''.()

0.4 &
β(z) = 0.106 + 0.118 × sin ( * 5.142. …(2c)
0.3 ''./0
0.2
The magnitude scaling factor (MSF) is used to
0.1
account for duration effects (i.e., number of
0
loading cycles) on the triggering of liquefaction.
0 5 10 15 The MSF factor is applied to the calculated value
Period (sec) of CSR for an earthquake of magnitude Mw =7.5
through introduction of magnitude scaling factor
Fig.7 Response spectra at the ground surface (MSF). MSF accounts for the duration effect of
for BH 2 Tollygunge ground motions. MSF for Mw < 7.5 is expressed as
follows (Idriss and Boulanger 2014):
INDIAN GEOTECHNICAL SOCIETY, KOLKATA CHAPTER
GEOTECHNICS FOR INFRASTRUCTURE DEVELOPMENT

KOLKATA 11th – 12th March 2016, Kolkata, West Bengal, India

34 Table 4 Rod Length Correction factors


MSF=1+ (MSFmax-1)(8.64 exp( . 6 1.325. ..(3a)
5
Rod Length (m) CR
78 . <3 0.75
MSFmax = 1.09+ ( 9 :;<= .2 2.2……….(3b)
)>
3-4 0.80
Since the liquefaction resistance increases with 4-6 0.85
increasing confining stress, the overburden 6-10 0.95
correction factor (K? ) is applied such that the 10-30 1.00
values of CSR are adjusted to an equivalent
overburden pressure σv0 of 1 atmosphere. The K? Table 5 Sampler Correction factors
relationship developed by Boulanger (2003) is as
follows: Sampling Method Cs

Standard Sampler 1.00


K? = 1- C? ln 7 . 1.1…………..74a.
C
Sampler without liner 1.10-1.30
'
C? = G 0.3 ...............(4b)
'0.E3/.>>F789 .:;<=
Table 6 Borehole Diameter correction
Factors
SPT-N value Correction Factors Bore Hole Diameter (mm) CB
Before the calculation of Cyclic Resistance Raio 65-115 1.00
(CRR) the N values obtained from the standard 150 1.05
penetration test must be corrected for the following
200 1.15
factors: overburden, rod length, non-standard
sampler, borehole diameter and hammer energy
Determination of Cyclic Resistance Ratio
efficiency; resulting in standardized SPT value
The Cyclic Resistance Ratio (CRR) of soil is
(N1,60) to a specific energy level. The relation
usually correlated to an in-situ parameter SPT blow
between SPT value (N) and the correction factors
count or CPT resistance or shear wave velocity
are illustrated below.
(Vs). Here, SPT based corrrelation is given as
(N1)60 = corrected SPT blow count to 60% energy
follows:
level = N×CN×CE×CB×CR×CS ………….(5) 789 .:;<=
CRRM=7.5,σv ′ =1 = exp[ +
'5.'
Where, CN is a factor to normalise N to a common
reference effective overburden stress; CE is 789 .:;<= 2 78 . 78 .
hammer energy efficicency factor; CB is borehole ( ) - ( 9 :;<= )3 + ( 9 :;<= )4 -2.8]…(6)
'/K /).K />.5
diameter correction factor; CR is rod length
correction factor and CS is correction for samplers (N1)60cs = (N1)60+ ∆ (N1)60 ……….(7a)
with or without liners.CE has been taken as 1 in this
case. E.( '>.(
∆(N1)60 = exp(1.63 + –( )2 )….(7b)
OPQL.L' OPQL.L'
I
CN = ( )m G 1.7 where, ∆(N1)60 is the equivalent clean sand
adjustment; ∆(N1)60cs is the equivalent clean sand
m= 0.784-0.0768xF7J' .KLMN corrected SPT value and FC is the fines content in
%.
7J' .KLMN value should be limited to 46.
Tanumaya Mitra, Ambarish Ghosh

Factor of Safety against Liquefaction where z is depth of the midpoint of the soil layer (0
Factor of safety against liquefaction is given as the to 20 m) and dz is differential increment of depth.
ratio of the cyclic resistance ratio to the cyclic The weighting factor, w(z), and the severity factor,
stress ratio generated during earthquake. F(z), are calculated as per the following
expressions:
PRRSTU.V,X T9 Y 4ZO Y [X
FOS = ……………(8)
PZR F(z) = 1 – FS for FS ≤ 1, and
Assessment of liquefaction potential index
F(z) = 0 for FS > 1, and w(z) = 10 – 0.5 z.
The liquefaction potential index (LPI) quantifies
the severity of liquefaction and predicts surface
According to Luna and Frost 1998), for the soil
manifestations of liquefaction, liquefaction damage
profiles with the depth less than 20 m, LPI is
or failure potential of a liquefaction-prone area
calculated using the following expression:
(Luna and Frost, 1998). Liquefaction Potential
index (LPI) has been determined by performing
LPI = ∑dbe' _b ]b cb ………………….(10)
weighted integration of 1-FOS values at each site
(Iwasaki et al., 1978; 1982) along the entire depth
With
of soil column limited to the depths ranging from 0
Fi = 1-FSi for FSi <1.0
to 20 m below the ground surface at a specific
location. The level of liquefaction severity with Fi= 0 for FSi 1.0
respect to LPI as per Iwasaki et al. (1982), Luna
and Frost (1998), and MERM (2003) is given in Where, Hi is thickness of the discretized soil
Table 7. Value of LPI can range from 0 for a site layers; n is the number of layers; Fi is liquefaction
with no liquefaction potential to a maximum 0f 100 severity for i-th layer; Fsi is the factor of safety for
for a site in which the factor of safety is zero over i-th layer; wi is the weighting factor(=10-0.5 Zi);
the entire 20 m depth range. and Zi is the depth of i-th layer (m).

Table 7 LPI values RESULTS

Luna and CSR AND CRR


Iwasaki et Frost Factor of Safety
LPI al.(1982) (1998) MERM(2003) 0 1 2
0 10
0
Little to CSR 0
5
LPI=0 Very low none None 5
10 CRR
10
Depth (m)

15 15
Depth (m)

0<LPI<5 low Minor Low 20 20


25 25
30 30
5<LPI<15 High Moderate Medium
35 35
40 40
15<LPI Very high Major High 45
45
50
50

Iwasaki et al.,(1978, 1982) proposed that the Fig.8 Liquefaction Analysis BH 1-Dumdum
liquefaction potential index (LPI) is expressed as PGA 0.178 g
follows:
/L
LPI = \L ]7^. _7^.`^ .................(9)
INDIAN GEOTECHNICAL SOCIETY, KOLKATA CHAPTER
GEOTECHNICS FOR INFRASTRUCTURE DEVELOPMENT

KOLKATA 11th – 12th March 2016, Kolkata, West Bengal, India

CSR and CRR Factor of Safety Factor of Safety Factor of Safety


0 2 0 10 0 10 0 10
0 0 0 0
CSR
5 5 5 PGA 0.178 g PGA 0.151 g
10 CRR 10 10 10
Depth (m)

PGA 0.205 g

Depth (m)
15 15 PGA 0.255 g
15
Depth (m)
20

Depth (m)
20 20 20
25 25 25 30
30 30 30
35 35 40
35
40 40
40
45 45 50
45
50
50 50 DUMDUM 60 TOLLYGUNGE

Fig.9 Liquefaction Analysis BH 1-Dumdum Fig.12 Comparison of Factor of Safety for


PGA 0.255 g different PGA

CSR and CRR Factor of Safety Depth of


BH
Site PGA Liquefaction LPI Type
0 2 No. (m)
0 10
0 0.178g 18 41.61
0 Dumdum 1
CSR
10 0.255g 18 55.88 Very High
CRR 10 Liquefiable
0.151g 16.5 37.94
20 Tollygunge 2
Depth (m)

20 0.205g 16.5
Depth (m)

50.96
30 Table 8 Summary of Liquefaction analysis
30
40 CONCLUSION
40
50 Seismic ground response analysis have been
50 carried out at two borehole locations with two
60 60 different PBRA values of 0.10g and 0.16 g using
equivalent linear approach in DEEPSOIL software.
Fig.10 Liquefaction Analysis The PGA values so obtained have been used for
BH 2-Tollygunge PGA 0.151 g liquefaction analysis in a code written in
CSR and CRR Factor of Safety MATLAB and the results have been shown in a
0 2 0 10
graphical format; the intersection of CSR and CRR
0
will give the probable extent of liquefaction (i.e.,
0
CSR FOS<1); From the above results it is evident that
10 CRR 10 the locations analyzed are very highly susceptible
to liquefaction as per the LPI values.
20 20
Depth (m)

Depth (m)

30 30 REFERENCES

40 40 1.Hashash, Y.M.A., Musgrove, M.I., Harmon,


J.A., Groholski, D.R., Phillips, C.A., and Park,
50 50
D. (2015) “DEEPSOIL 6.1, User Manual”.
60 60

Fig.11 Liquefaction Analysis


BH 2-Tollygunge PGA 0.205 g
Tanumaya Mitra, Ambarish Ghosh

2.Idriss, I.M and Boulanger, R.W (2004)


“Evaluating the potential for liquefaction or
cyclic failure of silts and clays.” Centre for
Geotechnical Modelling, University of California
at Berkeley, Report No. UCD/CGM-04/01.
3.Idriss, I.M and Boulanger, R.W (2010) “SPT-
Based Liquefaction Trigerring Procedure”.Centre
for Geotechnical Modelling, University of
California at Berkeley, Report No.UCD/CGM-
10-02.
4.Idriss, I.M and Boulanger, R.W (2014) “CPT
And SPT Based Liquefaction Triggering
Procedure”. Centre for Geotechnical Modelling,
University of California at Berkeley, Report No
UCD/CGM-14/01.
5.Iwasaki, T., Tokida, K., Tatsuko, F., and Yasuda,
S.: A practical method for assessing soil
liquefaction potential based on case studies at
various sites in Japan, Proceedings of 2nd
International Conference on Microzonation, San
Francisco, 885–896, 1978.
6.Kramer, Steven L. 1996. Geotechnical
Earthquake Engineering. Prentice Hall.
7.Luna, R. and Frost, J. D.: Spatial liquefaction
analysis system, J. Comput. Civil Eng., 12, 48–
56, 1998.
8. Nath, S.K, Adhikary, M.D, Maiti, S.K , Devaraj,
N, Srivastava, N and Mohapatra, L.D (2014)
“Earthquake scenario in West Bengal with
emphasis on seismic hazard microzonation of the
city of Kolkata, India”. Nat. Hazards Earth Syst.
Sci., 14, 2549– 2575, 2014.
9.Youd, T.L., Idriss, I.M.,Andrus, R.D.(2001)
Liquefaction Resistance of soils: Summary
Report from the 1996 NCEER and 1998
NCEER/NSF Workshops on Evaluation of
Liquefaction Resistance of Soils.

Potrebbero piacerti anche