Liquefaction Assessment With GeoStudio
Liquefaction Assessment With GeoStudio
Liquefaction Assessment With GeoStudio
com
The deviator stress q represents the shear in the soil sample. In a triaxial test setup, q is equal to (σ1 - σ3)
where σ1 is the major principal stress and σ3 is the minor principal stress.
The parameter p´ is the mean effective stress, which is defined in terms of effective principal stress as
follows:
p
1 2 3
3
Page 1 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
In a triaxial test, σ2 is equal to σ3 and the mean effective stress then becomes,
p
1 2 3
3
The critical state line (CSL) represents the strength developed at large strains when the shear resistance
and volume remain constant with continued ongoing strain. The critical state strength is sometimes also
referred to as the steady-state strength. Fundamentally, the critical-state and steady-state definitions are
slightly different but for practical purposes here the two can be considered to be analogous and will
consequently be used interchangeably in this discussion.
The slope of the CSL in a q-p´ stress space is usually defined by the capital Greek letter Μ (mu). The
slope Μ is related to the effective friction angle ´ by:
6sin
3 sin
q
Collapse point
Steady-state
strength
A p'
Figure 2 Effective stress path for loose sand in an undrained triaxial test
If a series of undrained triaxial tests are completed on a series of samples at the same initial void ratio the
stress paths will appear as illustrated in Figure 3. A straight line can be drawn from the steady-state
strength through the peaks or collapse points. Sladen, D’Hollander and Krahn (1985a) called this line a
Collapse Surface.
Page 2 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Collapse surface
Steady-state
strength
p'
Figure 3 Collapse surface illustration
Some researchers have proposed that the collapse surface line should pass through the origin of the plot
(e.g., Vaid and Chern (1983). Since flow liquefaction cannot occur below the steady-state point, the
authors crop the lower end of the collapse line as illustrated in Figure 4.
As noted, a soil can liquefy when the stress path reaches the collapse surface during undrained loading;
consequently, Kramer (1996, p. 363) named the collapse surface a flow liquefaction surface (FSL). As
discussed subsequently, other research has demonstrated that loose sands can also collapse under drained
loading; however, once collapse is initiated the soil can either: a) liquefy and follow a similar stress path
as shown in Figure 2 if drainage is impeded; or b) undergo volumetric compression if drainage is
permitted, allowing the stress path to proceed more steadily towards the CSL. Research has shown that
collapse can be initiated during drained loading for both dry and saturated soils. The FSL designation is
consequently too narrow because a soil can collapse without ‘flowing’ (that is, liquefying). The collapse
surface designation is more general.
The undrained steady-state strength of loose sand tends to be a relatively small value. Consequently, the
difference between whether the collapse surface is projected through the origin or through the steady-state
point on the CSL is relatively small. For practical purposes and the interpretation of field behavior the
two approaches are in essence the same.
Page 3 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Chu et al. (2003) called the collapse line an Instability Line (IL). The reasons for this choice of
nomenclature is rather perplexing considering that the collapse phenomenon had been adequately
described by previous researches. Nonetheless, the instability line is essentially equivalent to the FLS and
the collapse line.
The more general collapse surface designation and a line that passes through the steady-state point on the
CSL are used here and in GeoStudio.
Drained conditions
Sasitharan et al. (1993) at the University of Alberta were able to clearly demonstrate that the sand grain-
structure can collapse during fully drained loading as well as during undrained loading. Collapse can
occur at a mobilized friction angle m that is well below the conventional effective friction angle .
The fact that collapse of a loose sand grain-structure can be initiated during fully drained loading is
critical to understanding the stability of the sand slopes. The initiation of collapse during drained loading
can result in: a) an undrained liquefaction type response resulting in rapid loss of strength (described
previously); or b) a drained response that is characterized by a decrease in the void ratio (that is, collapse).
It is tempting to conclude that scenario (b) would improve the stability of a slope because the decrease in
void ratio leads to a more stable grain-structure. This, however, is not necessarily the case because
collapse in one region of a slope would cause the redistribution of stresses to other regions of the slope.
This would in turn promote a slope failure.
Dry sand
Skopek et al. (1994) at the University of Alberta demonstrated that collapse of a loose sand grain-
structure can even occur in dry sand (Figure 5 and Figure 6). The soil specimens were loaded by
decreasing the mean effective stress at a constant deviator stress.
Of great significance is the associated change in volume or void ratio. Initially, the void ratio remained
relatively constant while the mean effective stress diminished (Figure 6), but then the void ratio suddenly
decreased dramatically, particularly for the very loose sample. The sudden decrease in void ratio reflects
the collapse in the grain-structure. After the collapse, the void ratio continues to decrease with a further
reduction in the mean effective stress. This continues until the stress state reaches the CSL.
The tendency for volume change would have resulted in a dramatic and sudden increase in the pore
pressures had the specimen been initially saturated and the drainage valve closed at the point of collapse.
Such a condition would have triggered liquefaction.
Page 4 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
300
250
e
Lin
Shear stress q ( kPa )
a te
200 St
dy
ea
St
150 Loose
100
Very loose
50
0
0 50 100 150 200 250 300
Effective normal stress p'(kPa)
0.90
0.88
0.86
Collapse point
0.84
Very loose
Void ratio e
0.82
0.80
0.78
Loose
0.76
Stead y
0.74 State L
ine
0.72
0 50 100 150 200 250 300
Effective normal stress ( kPa )
Page 5 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
q
Collapse surface
Steady-state
strength
B
p'
Figure 7 Cyclic stress path from B to the collapse surface
Page 6 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Figure 8 Illustrative stress paths from X to collapse surface for grain-structure collapse
6sin L
L
3 sin L
The collapse surface inclination generally increases as the initial density increases. The lowest L values
correspond with very loose sand. As the density increases the inclination increases and may approach the
critical state line when the sand reaches a medium density. Chu et al. (2003) show collapse line slopes
greater than the CSL for dense sands but this issue will not be addressed in this report.
Figure 9 is a graph published by Chu et al. (2003). The inclinations L vary between 18.3 and 34.5
degrees (ML is between 0. 7 and 1.4). The corresponding void ratios vary between 0.972 and 0.864.
Page 7 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Figure 9 Collapse line inclination as a function of density (after Chu, et al. 2003)
Sladen et al. (1985a) presented the data for three different isotropically consolidated sands as shown in
Table 1. The lowest L is 14.3 degrees and the highest is 18.5 degrees.
Table 1 Table of data showing measured collapse line inclinations (after Sladen et al. 1985b)
Lade (1993) presented a summary of L values as shown in Figure 10. Despite the scatter, there is clear
evidence for the trend of increasing L with increasing density. For relative densities less than 50 percent
the range is about 15 to 25 degrees.
Page 8 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Kramer (1996, p. 364) makes this statement in his text book: for ‘isotropic initial conditions, the slope of
the FLS ( L ) is often about two-thirds the slope of the drained failure envelope for clean sand.’ These
data can be used as a guide for selecting L values.
Steady-state strengths
The steady-state strengths of loose sand tend to be relatively small. Sladen et al. (1985b) were of the
view that the undrained steady-state strengths at the Nerlerk Berm failures in the Beaufort Sea were less
than about 2 kPa. This was based on back analyses of the failures and the eventual very flat slopes of the
sliding mass.
The data presented in Figure 9, also suggests that the steady-state strength for loose clean sands is very
low as the curves tend to converge near the origin of the graph.
Castro el al. (1992), as part of a re-evaluation of the liquefaction failure that occurred at the Lower San
Fernando Dam in 1971, concluded that the undrained steady-state strength for the hydraulic fill in the
dam was likely in the range of 20 to 30 kPa. This relatively high value is likely reflective of the
significant fines content in the hydraulic fill.
The published information on the steady-state strength is rather meager and not sufficient to select a value
for stability analyses with any confidence. Fortunately, stability analyses are not all that sensitive to the
steady-state strength parameter. The analyses are much more sensitive to the inclination of the collapse
surface, which can be estimated more accurately.
Page 9 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
than for dirty sand, all else being equal”. They go on to state: “a high fines content will also reduce
permeability and increase compressibility making an undrained response to any given loading condition
more likely”.
More recently, Seid-Karbasi and Byrne (2007) have investigated the effect of silty-clayey layers on the
liquefaction behavior of sands. They have shown that these types of layers can act as barrier to the
dissipation of excess pore-pressure associated with the collapse of the sand grain structure and thereby
contribute to the potential instability. The inference is that in the absence of such impeding layers the
excess pore pressure could likely dissipate faster and the sand would fail more in a drained manner than in
an undrained manner. The outwash sand can be stratified with layers of varying fines content. This
stratification can played an important role in the behavior of the sand and the resulting stability of earth
structures.
Nerlerk Berm
In the early 1980’s artificial sand islands were constructed in the Beaufort Sea to facilitate hydrocarbon
exploration. Construction of the Nerlerk berm started in 1982 at a location where the sea depth was about
45 m. The berm was initially constructed by dumping sand from barges. More sand was later added on
the berm via hydraulic methods. The sand was dredged from the seabed and pumped through a floating
pipeline.
The sand placement caused at least five side-slope failures as illustrated in Figure 11. Subsequent studies
by Sladen et al. (1985a, 1985b) provided the collapse surface rational for the slope failures. Ultimately, it
was concluded that the failures were the result of liquefaction in the sand due to the static loading
resulting from the hydraulically placed sand.
Page 10 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Figure 11 Beaufort Sea Nerlerk Berm liquefaction slope failures (after Sladen et al. 1985b)
Page 11 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Figure 12 Profile of the Cougar 7 dump failure at the Greenhills Mine (after Dawson et al, 1998)
Figure 13 Plan view of run-out at the Cougar 7 dump failure at the Greenhills Mine (after Dawson
et al 1998)
Coal stockpiles
Dramatic flow-like slope failures have occurred in stockpiles of coking coal at a north Australian coal
export terminal. Some of the slips have flowed up to 60 m beyond the original stockpile toe when the
dumps where 10 to 14 m high (Eckersley, 1990). The failures occurred after the coal became saturated
(or nearly saturated) due to heavy rainfall. Eckersley (1990) also concluded that the failures were
initiated under essentially static drained conditions at a mobilized strength much less than what would be
represented by conventional peak effective stress strength parameters c´ and . The resulting initial
movements lead to the rapid generation of excess pore-pressures and the accompanying strength loss that
Page 12 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
caused the sudden acceleration of the sliding mass. Eckersley’s explanation of why the failures occurred
is entirely consistent with the concepts associated with a sand grain-structure collapse.
Under undrained conditions, the pore-pressures can rise sharply at the point of collapse and the strength
fall down suddenly to low undrained steady state strength. This sudden rise in pore-pressure and
associated strength loss can manifest itself in liquefaction.
The concept of a collapse surface is a highly useful tool for assessing the liquefaction potential of earth
structures.
Page 13 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Any stress state points in GeoStudio based on static stress that fall in this region are not marked as
liquefied prior to an earthquake analysis but may be marked as liquefied during the earthquake shaking as
the pore-pressures increase.
Any stress state points that fall into this region are of the greatest interest as far as assessing the potential
for liquefaction.
State points that fall in this region are not marked as liquefied or liquefiable in GeoStudio.
Page 14 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Numerically, in a GeoStudio analysis a stress state point can exist above the collapse surface. This is
results from an inaccurate description of the stress - distribution or collapse surface definition.
Practically what it means is that the soil is in a very precarious unstable state and that any amount of static
or dynamic disturbance could cause the sand grain-structure to collapse.
Page 15 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
In GeoStudio any computed stress state points that fall into this zone are marked as liquefied and no
excess pore-pressures are allowed to develop.
Now consider Points A and B at different initial shear stresses but at the same confining stress as
illustrated in Figure 18. Point A is again very close to the collapse surface and any amount of shaking
disturbance could cause liquefaction. Whereas Point B is a long way away from the collapse surface and
high excess pore-pressures would need to develop during earthquake shaking for liquefaction to occur.
Page 16 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
This illustrates why the initial static stress conditions are so critical to assessing the liquefaction potential.
Researchers that have used a cyclic stress approach to assessing the potential for liquefaction early on
recognized the strong influence of the initial static confining stress and initial static shear stress. From
this evolved what are known as Confining Stress (Ks) and Shear Stress (Ka) correction factors. Such
correction factors are not required in the context of a collapse surface.
More details on these correction factors are given in the QUAKE/W Engineering Book on pages 99 to
104.
The beauty about the collapse surface concept is that it inherently accounts for the initial shear and confining
stresses as illustrated above. No correction factors are required as in the cyclic stress approach.
Page 17 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
20
18
16
14
12
Elevation - m
10
The water table is at the ground surface. The total unit weight of the soil is 20 kN/m^3 and the unit weight
of the water is taken to be 10 kN/m^3 for convenient discussion purposes.
The convention effective strength parameters c´ and φ´ are zero and 30 degrees.
To begin with we’ll set Ko to 0.5.
The steady-state shear strength is set to 5 kPa; this makes qss = to 10 kPa.
Also, to start with we’ll make the collapse surface inclination 18 degrees.
Now we can do an Insitu type of analysis in SIGMA/W. It is necessary to select the Elastic-Plastic
Material Model for this analysis. The specified material properties are shown in the following screen
capture.
Page 18 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
The SIGMA/W Insitu results indicate that the lower 14 m of the soil column is liquefied or in a
liquefiable state as shown in Figure 20.
20
18
16
14
12
Elevation - m
10
Page 19 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
The reason for this is that the q/p´ ratio is above or very close to the specified collapse surface. This can
be vividly illustrated by taking the q and p´ stress profiles into EXCEL and plot these values relative to
the CSL and the position of the collapse surface. The end result is shown in Figure 21. At low stress
levels the field q/p´ ratio is just below the collapse surface and at higher stress levels the q/p´ ratio is just
above collapse surface and consequently the liquefaction shading in Figure 20.
180
160
140
Deviatoric Stress ‐ q
120
100
80
60
40
20
0
0 20 40 60 80 100 120 140 160 180 200
Mean effective stress ‐ p'
Figure 21 Field q/p´ ratios relative to the collapse surface with Ko equal to 0.5
If this was indeed representative of the actual field conditions, there would be little or no value doing a
QUAKE/W dynamic earthquake analysis. The simply SIGMA/W Insitu analysis implies that any amount
of strong motion shaking could in all likelyhood cause the soil to liquefy.
The situations for Ko equal to 0.6, 0.8 and 0.95 are shown in the following graphs.
When Ko is 0.6, some strong ground motion may result in sufficient generation of excess pore-pressure
for a significant portion of the column to liquefy. If Ko is 0.8 the field q/p´ is fairly far away from the
collapse surface and it would mean that large excess pore-pressures would need to develop for
liquefaction to occur. If Ko were to be 0.95, large excess pore-pressures might develop but liquefaction
would be highly unlikely.
The SIGMA/W Insitu analyses indicate no liquefaction shading when Ko is greater than 0.6. This is
because the field q/p´ ratios are not on or above the collapse surface.
Page 20 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
180
160
140
Deviatoric Stress ‐ q
120
100
80
60
40
20
0
0 20 40 60 80 100 120 140 160 180 200
Mean effective stress ‐ p'
Figure 22 Field q/p´ ratios relative to the collapse surface with Ko equal to 0.6
180
160
140
Deviatoric Stress ‐ q
120
100
80
60
40
20
0
0 20 40 60 80 100 120 140 160 180 200
Mean effective stress ‐ p'
Figure 23 Field q/p´ ratios relative to the collapse surface with Ko equal to 0.8
Page 21 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
180
160
140
Deviatoric Stress ‐ q
120
100
80
60
40
20
0
0 20 40 60 80 100 120 140 160 180 200
Mean effective stress ‐ p'
Figure 24 Field q/p´ ratios relative to the collapse surface with Ko equal to 0.95
From these simply SIGMA Insitu analyses we can readily see how strongly the potential for liquefaction
is influenced by the initial static insitu stress state conditions.
From these analyses we can see that potential for liquefaction is heavily influenced by the shear stresses
that exist in the ground prior to any earthquake shaking.
A simple preliminary SIGMA/W analysis can also help with understanding the generation of liquefaction
zones in a later QAUKE/W analysis. If the field q/p´ ratios are close to the collapse surface then
liquefactions zones will develop very quickly in a QUAKE/W analysis where as if the static field q/p´
ratios are far removed from the collapse surface the QUAKE/W analysis may not show any liquefaction
zones at the end of the earthquake shaking. This is discussed further below.
Page 22 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
6 Two-dimensional situation
Now let’s look at the following simple 2D case. It is assumed that the foundation material is loose and
potentially liquefiable. The dam embankment material is assigned the same material properties as the
foundation except that it is not deemed to be liquefiable; that is, no collapse properties are assigned to the
dam.
14
12
10
Underdrain
Elevation - m
0
0 5 10 15 20 25 30 35 40 45
Distance - m
Both materials are assigned an effective friction angle φ´ equal to 30 degrees with c´ equal to zero. This
makes the slope of the CSL equal to 1.2.
For illustrative purposes, the foundation is assigned collapse surface properties. The inclination is 18
degrees. The steady-state strength is set to zero. This is an unrealistic field value but it is useful for
illustration purposes. By making Css zero, the q/p´ ratio is a constant making it easier to compare the
liquefiable shaded areas with q/p´ contours. For 18 degrees the slope of the collapse surface in q - p´
space is 0.69.
The following figure shows the CSL and collapse surface with these parameters. Any q/p´ ratios that
fall between the two lines will be marked as liquefiable.
CSL Collapse surafce
200
180
160
140
Deviatoric Stress ‐ q
120
100
80
60
40
20
0
0 20 40 60 80 100 120 140 160 180 200
Mean effective stress ‐ p'
Page 23 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
We can do a SIGMA/W Insitu type of analysis to look at the situation under static insitu loading
conditions. The pore-pressure conditions will come from the Water Table definition. To obtain the
information of interest, it is necessary to use Effective-Drained Parameters with an Elastic-Plastic
constitutive model.
12
10
Underdrain
Elevation - m
0.7
1.2 1
2
0
0 5 10 15 20 25 30 35 40 45
Distance - m
12
10
Underdrain
Elevation - m
8
1
6 1.2
2 0.7
0
0 5 10 15 20 25 30 35 40 45
Distance - m
Page 24 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
12
10
Underdrain
Elevation - m
2
8 1. 1
0.7
4
0
0 5 10 15 20 25 30 35 40 45
Distance - m
12
10
Underdrain
Elevation - m
0
0 5 10 15 20 25 30 35 40 45
Distance - m
Figure 30 Liquefiable zone with Ko equal to 0.5 and Css equal to 5 kPa
Let us now make the assumption that the strength in the shaded area has fallen down to the undrained
steady-state strength for some undefined reason. We need to define a new material for the stability
Page 25 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
analysis. The new foundation material will have all the properties as before except the collapse surface
angle will be set to zero. This means that when the slip surface is in the shaded area the strength will be
5 kPa.
Consider the slip surface in the following figure. The cohesive and frictional strength along the slip
surface are shown in figure. Notice how in the middle portion of the slip surface the frictional strength is
zero and the cohesive strength is 5 kPa which represents the Css strength.
(Unfortunately, currently it is not possible to show the shaded liquefiable zone in SLOPE/W – watch for
this in a future version).
14
12
10
Underdrain
Elevation - m
0
0 5 10 15 20 25 30 35 40 45
Distance - m
25
20
15 Cohesion :
Cohesive
kPa
10
5
Friction :
Frictional
-5
10 12 14 16 18 20 22 24 26
X (m)
Page 26 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Often it is necessary to create a new material for the stability analysis if the objective is to look at the Css strength
alone; that is, use the Css strength but make the collapse surface inclination zero.
It is important to note at this stage that much can be done in the liquefaction evaluation process without
using QUAKE/W. Adding a QUAKE/W analysis is necessary only in the later stages of the evaluation
process, if at all. Sometimes, a definitive conclusion can be reached before even proceeding onto a
dynamic shaking analysis with QUAKE/W. If, for example, the margin of safety is already less than 1.0
under static conditions, then there is likely no value in doing a QUAKE/W analysis.
7 Case history
Now let’s look at the above discussion in the context of a case history. We can do this by abstracting
information from the QUAKE/W Detailed Example called the Upper San Fernando Dam.
Figure 33 Liquefiable zone in the Upper san Fernando Dam prior to the earthquake
What the initial conditions suggest is that there are potentially zones where the q/p´ ratio is very close to
the collapse surface and that any generation of excess pore-pressures may cause the stress state to move
onto the collapse surface. This is indeed what happens in this case as shown in Figure 34. The zones of
potential liquefaction have grown and developed.
In this case the QUAKE/W analysis is an essential part of the evaluation.
Page 27 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
1 El. 365
2
3
2.5
1
Figure 34 Liquefiable zones in the Upper San Fernando Dam after the earthquake
You will notice in this Detailed Example, that special materials have been created for the post-earthquake
analyses so that the liquefied zones have just the steady-state strength Css .
Page 28 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
Page 29 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
8 QUAKE/W analysis
It is very important to use only the strong motion portion of an earthquake time history record in the
QUAKE/W analysis step in a liquefaction evaluation. Including the early small trembling motion at the
start and the end of record makes the QUAKE/W analysis unnecessarily difficult (and even frustrating
sometimes). The computing time takes too long and it creates too much data making the viewing of
results too slow.
Remember, it is only the large dynamic shear stresses that cause the generation of excess pore-pressures
and this happens only during the most intense motion.
This is discussed on Pages 155 and 156 in the QUAKE/W Engineering Book.
The record shown in Figure 35 has a duration of almost 50 sec and includes over 12,000 data points. This
record could be easily reduced to the one in Figure 36 for a QUAKE/W liquefaction-assessment studying.
The modified record is only about 18 sec with about 5000 data points. The record could be further
modified by deleting every other data point without having a noticeable effect on the QUAKE/W results.
Modifying the earthquake record is often the most conveniently done in a spreadsheet (EXCEL). Once
the data has been imported into QUAKE/W, all the data can be selected and copied into EXCEL for
modification (right mouse click in the list box of the data). Once the record has been modified, you can
paste the data back into QUAKE/W through the clipboard.
The modification can also be done directly in QUAKE/W by group selection of certain portions of the
data points and clicking on delete.
0.6
0.4
0.2
Acceleration ( g )
-0.2
-0.4
-0.6
-0.8
0 10 20 30 40 50
Time (sec)
Page 30 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
0.6
0.4
0.2
Acceleration ( g )
-0.2
-0.4
-0.6
-0.8
0 5 10 15 20
Time (sec)
The first step is to remove the low trembling noise at the ends of the record and then consider removing
every other data point, for example, if the data was recorded at a very small time interval.
Remember, when in doubt you should do most of your preliminary work with a simplified analysis and
then near the end of the modeling try some more complicated analyses to determine if it makes a
significant difference to your conclusions.
9 Field Ko conditions
Earlier it was been demonstrated that the potential for liquefaction is highly dependent on the static insitu
stress state. This then begs the question, what is an appropriate Ko.
Generally, liquefaction is associated with loose fine sands. For a sand to be in a loose state in the field it
likely was deposited in a calm fluvial or sedimentation environment and has not been subject to past
loading and unloading. In a sense, it is like the material is normally consolidated.
For normally, consolidated soils Ko can be estimated from the relationship:
K o 1 sin
Looking at it another way, it is unlikely that loose liquefiable sands have a high Ko. If the sand has a high
Ko it is likely no longer in a loose state because the past loading and unloading that caused Ko to be high
also densified the soil.
These are general comments intended to start the GeoStudio user’s thought process on this issue. Clearly,
this is an important issue and needs to be assessed carefully in the context of project specific conditions.
Page 31 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
First and foremost, it is essential to have a clear understanding of the collapsible grain-
structure of loose fine sands. That the grain-structure can collapse and that the strength can
suddenly fall down to the undrained steady-state strength has been conclusively demonstrated
by laboratory tests and field observation.
The user must have an understanding of how GeoStudio flags elements as potentially
liquefiable in the context of regions in q-p´ stress space.
Recognize that the definition of a collapse surface automatically accounts (corrects) for the
shear and confining stress in the ground. No other correction factors are required like those
used in the cyclic stress approach.
Accept that fact that the liquefaction potential is tightly tied to the insitu static stress state.
Use SIGMA/W, and maybe SLOPE/W, to assess the situation before moving unto a
QUAKE/W analysis. A QUAKE/W analysis should only be undertaken after an initial
assessment based on static stresses. Starting with a complicated QUAKE/W analysis should
be avoided.
When a QUAKE/W analysis is undertaken, only the strong motion portion of a time history
record should be used.
To be clear that QUAKE/W alone cannot provide any information about permanent
deformations. QUAKE/W can only provide information about the dynamic inertial forces and
the resulting associated generation of excess pore-pressures.
Understand that permanent deformations can only be estimated by doing a SIGMA/W Stress
Redistribution type of analysis or a SIGMA/W Dynamic Deformation type of analysis.
Before doing a permanent deformation analysis, it is important to first check the stability
using the post-earthquake pore-pressures and reduced strengths resulting from collapse of the
sand-grain structure. If the stability analysis shows the structure to be unstable at this stage
(factor of safety close to or less than 1.0) then there is little value in a permanent deformation
analysis – the structure has already failed and collapsed. If such an analysis is done it may
provide a picture of the post-failure displacement field, as in the Lower San Fernando Dam
case history, but the magnitudes of the computed displacements will be meaningless. In
short, GeoStudio cannot be used for a post-failure deformation analysis.
Prepared by:
Dr. John Krahn
Page 32 of 33
GEO-SLOPE International Ltd, Calgary, Alberta, Canada www.geo-slope.com
11 References
Arya, L.M. and Paris, J.F. (1981). A physioempirical model to predict the soil moisture characteristic
from particle-size distribution and bulk density data. Soil Science Society of America Journal,
45: 1023-1030.
Castro, G., Seed, R.B., Keller, T.O. and Seed, H.B. (1992). Steady-state strength analysis of Lower San
Fernando dam slide, ASCE Journal of Geotechnical Engineering, Vol. 118, pp 406-427.
Chu, J., Leroueil, S. and Leong, W.K. (2003). Unstable behavior of sand and its implication for slope
stability, Canadian Geotechnical Journal, Vol. 40, pp 873-885.
Dawson, R.F., Morgenstern, N.R. and Stokes, A.W. (1998. Liquefaction flow slides in Rocky Mountain
coal mine waste dumps, Canadian Geotechnical Journal, Vol. 35, pp 328-343.
Eckersley, J.D. (1990). Instrumented laboratory flowslides, Geotechnique, Vol. 40, pp 489-502.
Gu, W.H. and Krahn, J., (2002). A Model for Soil Structure Mobility and Collapse, 15th ASCE
Engineering Mechanics Conference, June 2-5, 2002, Columbia University, New York, NY.
Krahn, J. 2003. The 2001 R.M. Hardy Lecture: The Limits of Limit Equilibrium Analyses. Canadian
Geotechnical Journal, Vol. 40, pp. 643-620.
Lade, P.V. (1993). Initiation of static instability in the submarine Nerlerk berm, Canadian Geotechnical
Journal, Vol. 30, pp 895-904.
Sasitharan, S., Robertson, P.K., Sego, D.C. and Morgenstern, N.R. (1993). Collapse behavior of sand,
Canadian Geotechnical Journal, Vol. 30, pp 569-577
Seid-Karbasi, M. and Byrne, P.M. (2007). Seismic liquefaction, lateral spreading and flow slides: a
numerical investigation into void redistribution, Canadian Geotechnical Journal, Vol. 44, pp 873-
890.
Skopek, P, Morgenstern, N.R., Robertson, P.K. and Sego, D.C., (1994). Collapse of Dry Sand, Canadian
Geotechnical Journal, Vol. 31, pp 1008-1014.
Sladen, J.A., D’Hollander, R.D. and Krahn, J., (1985a). The liquefaction of sands, a collapse surface
approach, Canadian Geotechnical Journal Vol.22, pp.564-578.
Sladen, J.A., D’Hollander, R.D., Krahn, J., and Mitchell, J.A., (1985b). Back analysis of the Nerlerk berm
liquefaction slides, Canadian Geotechnical Journal Vol.22, pp.579-588.
Vaid, Y.P. and Chern, J.C. (1983). The effect of static shear on resistance of liquefaction, Soils and
Foundations, Vol. 23, pp. 47-60.
Page 33 of 33