Creep In Soil Mechanics

0 10 20 30 40 50 60 70 80 90 Plasticity index, %

Figure 3.10. Undrained strength anisotropy from CK0U tests on NC clay (Jami-olkowski et al., 1985)

Figure 3.10. Undrained strength anisotropy from CK0U tests on NC clay (Jami-olkowski et al., 1985)

Swelling Pressure Test Soil Mechanism

Figure 3.11. Anisotropy of cu/ff'v for different soil at OCR=1 (Jardine and Menkiti 1999)

Another means to study anisotropy is through the Hollow Cylinder Apparatus (HCA) which has shown to be an excellent tool of anisotropy investigation as it allows the control of the principal stress orientation.

Jardine and Menkiti (1999) reported results of their experiences with the HCA on three silty soils (denoted by HRS, HK, KSS). Figure 3.11 reveals that the reduction of the undrained strength ratio (S) with the orientation of principal stresses (a = 0° for compression and a = 90° for extension) may be severe in some soils.

From figures 3.10 and 3.11 we can conclude the following:

• Less plastic, and often more sensitive, clays tend to have higher strength anisotropy than more plastic clays.

• The use of an undrained strength ratio estimated from compression tests (CK0U) in stability calculation will yield unsafe results for clays of low to moderate plasticity index and OCR.

It has been suggested (Jamiolkowski et al., 1985) that most stability analysis should consider strength anisotropy; at least the projects concerning staged construction or unusual loading. Unfortunately, anisotropy cannot be easily characterized and implemented in a model for use in routine work. Still, the assessment of TC and TE tests may help at selecting design parameters even when a simple model (e.g. isotropic) is employed.

3.7 Final recommendations

This section recommends procedures to obtain undrained strength ratio vs. OCR relationships required to calculate the initial cu profile and subsequent increases due to consolidation. The recommendations are divided into three levels of sophistication, as pointed out by Ladd (1991), depending on the degree of refinement required:

• Level A: For final design of all major projects and for sites where the foundation soils exhibit significant undrained stress-strain-strength anisotropy or contain unusual features (fissuring, varved, highly organic, etc.) and for projects requiring accurate predictions.

• Level B: For preliminary design and for final design of less important projects involving ordinary soils with low to moderate anisotropy.

• Level C: For preliminary feasibility studies and to check the reasonableness of initial strengths inferred from in-situ and laboratory conventional testing programs.

Levels A and B require laboratory CU testing to provide anisotropic and isotropic (average) input strengths, respectively, whereas level C relies on empirical correlations. Table 3.1 summarizes the testing programs recommended by Ladd (1991) from his experiences.

The CK0 U test for Level A use either the SHANSEP technique or simply Reconsolidation to the in-situ stress, depending on the soil type (e.g. highly structured), in-situ OCR and sample quality. The cu/a'v vs. OCR predictions, in the form of equation 3.2, should "exactly" simulate the in-situ response for the first stage of construction, but they may involve errors on the safe side when used to compute strength increases during consolidation.

For Level B programs, Ladd (1991) recommends the use of either CK0U direct simple shear or CK0 U triaxial compression and extension to estimate

Table 3.1. Recommended laboratory testing program (Ladd, 1991)

Level A

Level B

Level C

CK0U tests with different modes of failure: -Triaxial compression (TC) -Direct simple shear (DSS) -Triaxial extension (TE)

CK0U tests with either: -Direct simple shear (DSS) or

- Triaxial compression (TC) and Triaxial extension (TC) in order to estimate avrg. strength

Uses empirical correlations rather than testing. See section 4.5 for typical values.

a reasonable average value of shear strength along the potential failure surface of a slope. Moreover, he states that isotropic strength profiles suffice for the assessment of stability. Level B should not rely on tests performed on isotropically consolidated specimens. Level C should only rely on empirical correlations.

Levels A, B and C require a careful assessment of the stress history of the foundation soil. This fact, plus the observation that cu/a'v vs. OCR for most homogeneous soils falls within a fairly narrow range, means that consolidation testing usually represents the single most important experimental component for the design of staged construction projects.

Discussion on slope stability evaluation

W.F. Van Impe & R.D. Verastegui Flores Laboratory of Geotechnics, Ghent University, Belgium

4.1 Preamble

The objective of the present chapter is to review slope stability methods and related issues. The methods have been classified here into two categories: limit equilibrium methods and strength reduction methods.

The assessment of the stability of slopes remains a challenging task of geotechnical engineering. However, many aspects have been thoroughly studied over the last decades and today the methods of analysis are able to tackle complex problems.

Slopes, natural or man-made, are observed to collapse in different ways. Figure 4.1 summarizes some of the most common patterns of soil slope failures. Rocks and soft rocks slopes show different patterns out of the scope of this book.

The two major types of slides are rotational slides and translational slides (Fig. 4.1). Rotational slide are those in which the surface of sliding is curved concavely upward and the slide movement is roughly rotational about an axis parallel to the ground surface. On the other hand, a translational slide is one in which a soil mass moves along a roughly planar surface with little rotation. Such planar movement could be the result of the presence of a weak layer or an interface of different soil types.

Moreover, earthflows and creep are patterns observed especially in soft fine grained soils. Earthflows have a characteristic shape. They occur for example when the slope material liquefies and runs out forming a depression at the head and a mound at the toe. On the other hand, creep manifests as a imperceptibly slow, steady downward of the slope caused by for example, the environment action, presence of existing sliding surfaces and vicinity of stress state to failure.

From these 4 slope failure types, the rotational slide and translational slide were explicitly studied in classic soil mechanics by means of limit states methods. A short review is given in the next sections. Moreover, the stability analysis of quick clay masses is evaluated in detail.

Quick Clay

Earthflow Creep

Figure 4.1. Common patterns of soil slope failure (source USGS)

Earthflow Creep

Figure 4.1. Common patterns of soil slope failure (source USGS)

4.2 Causes of slope instability

When facing a design task it is important to understand the causes of instability of a slope to anticipate the changes in the properties of the soil that may occur over time, loading conditions, seepage conditions to which the slope will be subjected, etc.

As stated by Duncan and Wright (2005), when discussing the causes of slope failure it is useful to start from the very fundamental premise that the shear strength of the soil must be greater than the shear stress required at equilibrium. Consequently, the most fundamental cause of instability is that for some reason, the shear strength of the soil is less than the shear strength required for equilibrium and such condition can be reached in two ways:

• Through a decrease of shear strength in the soil

• Through an increase of the shear stress required for equilibrium

Figure 4.2. Cracking as an effect of environment action and its implication in slope stability
Figure 4.3. Results from fresh water leaching tests on clay specimens from Norway (Bjerrum, 1967)

Reasons for a decrease in the shear strength of the soil are for example an increase of pore water pressures (e.g. due to rainy seasons), cracking (i.e. due to the action of the environment and tension stresses, Fig. 4.2), swelling, leaching (Fig. 4.3), strain softening behavior, cyclic loading (e.g. leading to liquefaction). Figure 4.3 illustrates a striking example of soil behavior. These are results of leaching tests after Bjerrum (1967); they showed at that time how the structure of a marine deposited soil could significantly collapse when it is leached with fresh water creating sensitive clays. This particular example should always remind the engineer to stay alert and pay attention to the various ways of soil behavior. Special considerations on the analysis of quick clays are given in section §4.5.

Reasons for an increase of the shear stress required for equilibrium are for example an extra loading acting on the slope, water accumulation in cracks (Fig. 4.2), increase of the unit weight of the soil (e.g. due to wetting), excavation works at the toe of the slope, drop in water level at the site (e.g. due to water pumping), earthquake or other type of dynamic loading, etc.

In reality, slopes will fail usually because of a combination of some of the reasons cited above.

4.3 Stability conditions for analysis

The first requirement to perform slope stability analysis is to formulate correctly the problem. Selecting appropriate conditions for analysis of slopes requires considerations of the shear strength of soils under drained and undrained conditions, or under drainage conditions that will occur in the field.

The general principles involved in selecting analysis conditions and shear strengths are summarized in table 4.1.

When an embankment is constructed on a clay foundation, the embankment load causes the pore water pressure in the clay to increase. After a period of time, such increment will gradually dissipate and eventually the pore water pressures will return to the initial steady value. As the excess pore water pressure dissipates, the effective stresses in the foundation soil increase, the strength of the clay increases and as a result the factor of safety increases too. Figure 4.4 illustrates these relationships and out of it one may conclude that the most critical condition occurs at the end of construction (undrained). Then, it is only necessary to analyze the end-of-construction condition.

Table 4.1. Shear strength for stability analysis (Duncan, 1996)


End of construction Staged construction Long term

Procedure and Effective stress anal. Effective stress anal. Effective stress anal.

strength for with cC and ft with C and ft with C and ft sand

Procedure and Total stress anal. Total stress anal. Effective stress anal.

strength for with cu from approp. with C and ft clay consolidation anal.

When a slope in clay is created by excavation, the pore pressures in the clay decrease in response to the removal of the excavated material. Over time, the negative excess pore water pressure dissipate and the pore pressures eventually return to the initial steady value. As the pore water pressure increases, the effective stress in the decreases and the factor of safety decreases with time as illustrated in figure 4.4. Out of these relationships it can be concluded that the long-term (drained) condition is more critical than the end-of-construction condition.

Drained conditions are analyzed in terms of effective stresses using values of C and ft determined from drained tests, or from undrained tests with

Embankment Excavation

Embankment Excavation

Time Time

Figure 4.4. Stability conditions for analysis

Time Time

Figure 4.4. Stability conditions for analysis pore water pressure measurements. When dealing with clay, drained triaxial tests are frequently impractical because the required time is very long, therefore, undrained triaxial tests with pore pressure measurements are the most suitable. Values of ft for natural deposits of cohesionless soils are usually estimated using correlations from field tests results (i.e. CPT, SPT, DMT, PMT) given the current difficulties of testing high quality undisturbed sand samples.

Undrained conditions are analyzed in terms of total stress in order to avoid having to rely on estimated values of pore water pressure for undrained loading conditions. Undrained shear strength of soils is usually correlated from field tests or laboratory tests. For staged construction analysis, the undrained strength is furthermore assessed through consolidation analysis in combination with for example triaxial CU testing.

In cases where it is not clear whether the short-term or long-term condition will be more critical, both should be analyzed, to ensure that the slope will have an adequate stability under any condition.

Another important topic is that of the selection of an suitable factor of safety for design. The main considerations to take into account are the degree of uncertainty in evaluating conditions and shear strengths for analysis and the possible consequences of failure. Typical minimum acceptable values of factor of safety are about 1.3 for the categories end-of-construction and staged-construction and 1.5 for long-term conditions.

4.4 Stability analysis procedures

The universal availability of computers and a much improved understanding of the mechanics of slope stability analysis have brought about considerable changes in the computational aspects of slope stability analysis in the last years.

Single free-body approach: Infinite slope, Swedish circle.

Was this article helpful?

+1 0

Post a comment