NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

Module 3
DYNAMIC SOIL PROPERTIES
(Lectures 10 to 16)

Lecture 16
Topics
3.9 DAMPING RATIO
3.10 CYCLIC NONLINEAR MODELS
3.11 ADVANCED CONSTITUTIVE MODELS
3.12 STRENGTH OF CYCLICALLY LOADED SOILS
3.12.1 Definitions of failure
3.12.2 Cyclic strength
3.12.3 Monotonic strength

3.9 DAMPING RATIO

Theoretically, no hysteretic dissipation of energy takes place at strain below the
linear cyclic threshold shear strain. Experimental evidence, however, shows that
some energy is dissipated even at very low strain levels (the mechanism is not well
understood), so the damping ratio is never zero. above the threshold strain, the
breadth of the hysteresis loops exhibited by a cyclically loaded soil increase with
increasing cyclic strain amplitude, which indicates that the damping ratio increases
with increasing strain amplitude.

Damping ratios of highly plastic soils are lower than those of low plasticity soils at
the same cyclic strain amplitude (figure 3.50) The damping behavior of gravel is
very similar to that of sand (Seed et al., 1984).

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 1

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

Figure 3.50: Variation of damping ratio of fine-grained soil with cyclic shear
strain amplitude and plasticity index. (After Vucetic and Dobry 1991)

Damping behavior is also influenced by effective confining pressure, particularly

for soils of low plasticity, Ishibashi and Zhang (1993) developed an empirical
expression for the damping ratio of plastic and no plastic soils. Using (equation
3.52) to compute the modulus reduction factor, , the damping ratio is given
by

[ ( ) ] (3.54)

The influence of various environmental and loading conditions on the damping ratio
of normally consolidated and moderately over consolidated soils is described in
(table 3.9).

Table 3.8 Effect of Environmental and Loading Conditions on Modulus Ratio (at a
Given Strain Level) of Normally Consolidated and Moderately Over consolidated
Soils (from Dobry and Vucetic 1987).
Increasing Factor
Confining pressure, Increases with ; effect decreases
with increasing PI
Void ratio, Increases with
Geologic age, May increase with
Cementation, May increase with
Over consolidation ratio, OCR Not affected
Plasticity index, PI Increases with PI
Cyclic strain, Decreases with
Strain rate, G increases with but
/ probably not affected if G and

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 2

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

are measured at same

Number of loading cycles, N Decreases after N cycles of large
( measured before N cycles) for
clays; for sands, can increase (under
drained conditions) decrease (under
undrained conditions)

3.10 CYCLIC NONLINEAR MODELS

The nonlinear stress-strain behavior of soils can be represented more accurately by
cyclic non-linear models that follow the actual stress-strain path during cyclic
loading. Such models are able to represent the shear strength of the soil, and with an
appropriate pore pressure generation model, changes in effective stress during
undrained cyclic loading. A variety of cyclic nonlinear models have been
developed; all are characterized by (1) a backbone curve and (2) a series of rules
that govern unloading-reloading behavior, stiffness degradation, and other effects.
The simplest of these models have relatively simple backbone curves and only a few
basic rules. The applicability of cyclic nonlinear models, however, is generally
restricted to a fairly narrow, albeit important range of initial conditions and stress
paths.

The performance of cyclic nonlinear models can be illustrated by a very simple

example in which the shape of the backbone curve is described by . The
shape of any backbone curve is tied to two parameters, the initial (low-strain)
stiffness and the (high-strain) shear strength of the soil. For the simple example, the
backbone function, , can be described by a hyperbola

| |
(3.55)

Table 3.9 Effects of Environmental and Loading Conditions on Damping

Ratio of Normally and Moderately Overconsolidated Soils (from Dobry and
Vucetic 1987).
Increasing Factor
Confining pressure, Decreases with ; effect decreases
with increasing PI
Void ratio, Decreases with
Geologic age, Decrease with
Cementation, May decrease with
Overconsolidation ratio, OCR Not affected
Plasticity index, PI Decreases with PI
Cyclic strain, Increases with
Strain rate, Stays constant or may increase with
Number of loading cycles, N Not significant for moderate and N

The shape of the hyperbola backbone curve is illustrated in (figure 3.51). Other
expressions (e.g., the RAmberg-Osgood model (Ramberg and Osgood, 1943) can
also be used to describe the backbone curve. Alternatively, backbone curves can be

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 3

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

constructed from modulus reduction curves.

Figure 3.51: Hyperbolic backbone curve asymptotic to

The quantities may be measured directly, computed, or obtained by

empirical correlation. For the example model, the response of the soil to cyclic
loading is governed by the following four rules:

1. For initial loading, the stress-strain curve follows the backbone curve.
2. If a stress reversal occurs at a point defined by ( the stress-strain curve
follows a path given by

( )

In other words, the unloading and reloading curves have the same shape as the
backbone curve (with the origin shifted to the loading reversal point) but is
enlarged by a factor of 2. These first two rules, which describe Masing behavior
(Masing, 1926), are not sufficient to describe soil response under general cyclic
loading. As a result, additional rules are needed.

3. If the unloading or reloading curve exceeds the maximum past strain and
intersects the backbone curve, it follows the backbone curve until the next stress
reversal.
4. If an unloading or reloading curve crosses an unloading or reloading curve from
the previous cycle, the stress-strain curve follows that of the previous cycle.

Models that follow these four rules are often called extended Masing models. An
example of the extended Masing model is shown in (figure 3.52). Cyclic loading
begins at point A, and the stress-strain curve during initial loading (from A to B)
follows the backbone curve as required by rule 1. At point B, the loading is reversed
and the unloading portion of the stress-strain curve moves away from B along the
path required by rule 2. Note that the initial unloading modulus is equal to .

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 4

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

The unloading path intersects the backbone curve at point C, and according to rule
3, continues along the backbone curve until the next loading reversal at point D. the
reloading curve then moves away from D as required by rule 2, and the process is
repeated for the remainder of the applied loading. Although this model is very
simple and is expressed only in terms of effective stresses, it inherently incorporates
the hysteretic nature of damping and the strain-dependence of the shear modulus
and damping ratio. Other unloading-reloading models are available (e.g., Iwan,
1967; Finn et al., 1977; Vucetic, 1990); the Cundall-Pyke, 1979) is particularly
straight forward and easily implemented into ground response analyses. To avoid
spurious response at very low strain levels, some cyclic nonlinear models require
the addition of a small amount of low strain damping. Note that the cyclic nonlinear
model does not require the shear strain to be zero when the shear stress is zero. The
ability to represent the development of permanent strains is one of the most
important advantages of cyclic nonlinear models over equivalent linear models.

Figure 3.52: Extended Massing rules: (a) variation of shear stress with time;
(b) resulting stress-strain behavior (backbone curve indicated by dashed line)

This simple example model does not, however, allow for the determination of shear-
induced volumetric strains that can lead to hardening under drained conditions or to
pore pressure development with attendant stiffness degradation under undrained
conditions. Such factors are accounted for in the majority of the cyclic nonlinear
models commonly used in geotechnical earthquake engineering practice (e.g., Finn
et al., 1977; Pyke, 1979, 1985).

The ability to compute changes in pore pressure, hence also changes in effective
stress, represent another significant advantages of cyclic nonlinear models over
equivalent linear models. As pore pressures increase, effective stresses decrease,
and consequently the values of decrease. Since the shape and
position of the backbone curve depends on the backbone curve
degrades with increasing pore pressure. As with actual soils, the stiffness in a
stress-strain model depends not only on the cyclic strain amplitude, as implied by
the equivalent linear model, but also on the stress history of the soil.
ADVANCED CONSTITUTIVE MODELS
The most accurate and general methods for representation of soil behavior are based
on advanced constitutive models that use basic principles of mechanics to describe
observed soil behavior for (a) general initial stress conditions, (b) a wide variety of

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 5

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

stress paths, (c) rotating principal stress axes, (d) cyclic or monotonic loading, (e)
high or low strain rates, and (f) drained or undrained conditions.

Such models generally require a yield surface that describes the limiting stress
conditions for which elastic behavior is observed, a hardening law that describes
changes in the size and shape of the yield surface as plastic deformation occurs, and
a flow rule that relates increments of plastic strain to increment of stress. The Cam-
Clay (Roscoe and Schofield, 1963) and modified Cam-Clay (Roscoe and Burland,
1968) models were among the first of this type. Improvement in the prediction of
shear strains have resulted from the use of multiple nested yield loci within the yield
surface (Mroz, 1967; Pervost 1977) and the development of bounding surface
models (Dafalias and Popov, 1979) which incorporate a smooth transition from
elastic to plastic behavior. Detailed treatment of such advanced constitutive models
is beyond the scope of this book. The interested reader can refer to a number of
sources, including Desai and Siriwardane (1984), Defalias and Hermann (1982),
Wroth and Houlsley (1985), Lade (1988), and Wood (1991).

3.11 STRENGTH OF CYCLICALLY LOADED SOILS

The effect of cyclic loading on the limiting strength of soils is considerable
importance in geotechnical earthquake engineering. Problems of slope stability,
foundation performance, and retaining wall behavior, amount others, are strongly
influenced by the strength that the soil can mobilize at large strains.

3.12.1 Definitions of failure

The shear strength of an element of soil is typically defined as the shear stress
mobilized at the point of failure, but failure can be defined in many different
ways. In the field, failure is usually associated with deformations that exceed some
serviceability limit. Since deformation results from the integration of strains over
some volume of soil, the point of failure of an element of soil is often defined in
terms of a limiting strain.

Consider an element of soil in drained equilibrium under anisotropic stress

conditions in a cyclic direct simple shear test (point A in figure 3.53). The
application of a cyclic shear stress, , produces (under stress-controlled
conditions) a cyclic shear strain, , but also an increase in the average strain,
. The average shear strain increases with increasing numbers of loading cycles.
Clearly the strength of the soil during cyclic loading could be defined in terms of
limiting values of or or of some combination of the two. The available
strength of the sol under monotonic loading (after the cyclic loading has ended) may
also be of interest.

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 6

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

Figure 3.53: Definition of average and cyclic shear stress and shear strain
(After Goulois et al., 1985)

3.12.2 Cyclic strength

The levels of both cycles and permanent deformations are of interest in a number of
geotechnical earthquake engineering problems. They are also important in the
design of foundations for marine structures subjected to wave loading, and much of
the current state of knowledge of cyclic strength has come from research in that
area. The cyclic strength of an element of soil depends on the relationship between
the average shear stress, , and the cyclic shear stress, . When the average
shear stress is low, unidirectional strains will accumulate slowly, so the average
shear strain will also be low. The amplitude of the cyclic strain, however, may
become large if the cyclic shear stress is large. If, on the other hand, the average
shear stress is high (relative to the static shear strength, ), substantial
unidirectional strains can develop even when the cyclic shear stress is small.

For the case of , no unidirectional strain will develop, so failure must be

defined in terms of the cyclic shear strain, . When failure is defined in terms of
a specific level of cyclic shear strain (often 3 percent), the cyclic strength ratio,
defined as , decreases with increasing numbers of cycles as shown in (figure
3.54).

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 7

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

Figure 3.54: Variation of cyclic strength ratio with number of cycles for
different soils. (After Lee and Focht, 1976)

For cases in which is greater than zero, both , will depend on

and (Seed and Chan, 1966). Investigations of the cyclic response of marine
clays (e.g., Meimon and Hicher, 1980; Goulois et al. 1985; Anderson et al., 1988)
have shown that depends predominantly on and the number of cycles, and
depends predominantly on and the numbers of cycles (figure 3.55).

Figure 3.55: variation of average shear strain with average shear stress, cyclic
shear stress, and number of cycles in cyclic direct simple shear tests on plastic
drammen clay. (After Goulois et al., 1985.)

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 8

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

3.12.3 Monotonic strength

Evaluation of the static stability of slopes and retaining walls and the capacity of
foundations after earthquake shaking has ended is another important problem in
geotechnical earthquake engineering. Such problems require evaluation of the
available shear strength of the soil after the earthquake has ended. This post
earthquake strength must reflect any effects of cyclic loading imposed by the
earthquake.

The ultimate (residual, high-strain) undrained shear strength of a saturated soil is

controlled by its void ratio and structure. Barring any change in soil structure, a
saturated soil at a particular void ratio will mobilize a specific undrained strength,
with little influence of the history of stresses and strains by which that strength is
arrived at. For such soil conditions, the undrained strength after cyclic loading will
be equal to the undrained strength before undrained loading (at the same strain rate).
Since cyclic loading induces positive excess pore pressures, the effective stress in an
element of soil sheared monotonically after being subjected to cyclic loading will be
lower than that in an identical element that is sheared monotonically without prior
cyclic loading. Consequently, the element that had been cycled would be expected
to exhibit more delative behavior but to have a lower stiffness in the early stages of
monotonic undrained loading than the element that had not been cycled.

Changes in monotonic strength can be caused by disturbance of the soil structure

during cyclic loading. The extent to which the structure of the soil is disturbed is
influenced by the relationship between the cyclic strain amplitude and the strain at
which failure occurs under monotonic loading conditions (Thiers and Seed, 1969).
Substantial structural disturbance can modify the stress-strain behavior and reduce
the monotonic shear strength. The six triaxial specimens shown in (figure 3.56) had
similar void ratios (except specimen 6, which had a somewhat higher void ratio than
the rest) at the end of consolidation. Specimen 1 was sheared monotonically
immediately after consolidation, but specimens 2 to 6 were first subjected to varying
levels of cyclic loading. Since the void ratios were nearly the same, the specimens
would therefore be expected to have similar monotonic strengths. As shown by the
stress-strain curves and stress paths, they behaved largely as would be expected.
After being subjected to different levels of cyclic strain, their ultimate (large strain)
strength was similar (except specimen 6, which was lower than the others).
Differences in the ultimate strength can be explained by small differences in the
void ratios and also by differences in the extent of structural disturbance induced by
the cyclic loading.

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 9

NPTEL- GEOTECHNICAL EARTHQUAKE ENGINEERING

Figure 3.56: Effect of cyclic loading on subsequent monotonic undrained

loading behavior of traixial specimens of slightly plastic silt: (a) stress-strain
behavior; (b) effective stress path behavior. (After Castro and of Christian,
1976)

Thiers and Seed (1969) found that the ultimate strength of three clays decreased by
less than 10% when the cyclic strain amplitude was less than one-half of the failure
strain from monotonic tests. At higher cyclic strain amplitudes, the reduction in
strength was more dramatic, as illustrated in (figure 3.57). Similar results have been
obtained by others (e.g., Koutsoftas, 1978; Byrne et al., 1984).

Figure 3.57: Effect of peak cyclic strain on monotonic strength after cyclic
loading. ( Thiers and Seed, (1969))

Dept. of Civil Engg. Indian Institute of Technology, Kanpur 10