CREEP CHARACTERIZATION OF TROPICAL SOILS
H.G. Brandes(P.I.), D.D. Nakayama, C.S.L. Tsui
Funded By: Hawaii Department of Transportation and the Federal Highways Administration
Department of Civil Engineering
Content
Laboratory
Testing Program
Introduction
Creep is defined as a continuing
deformation under a constant state of stress. It is of particular concern in fine-grained soils, where
creep and creep failure can occur under loads significantly less than design
standards. In principal, all real
bodies, including soils, will deform over a period of time. However, these time-effects are often
neglected in engineering practice, often with disastrous consequences. Creep testing includes the
determination of maximum applied loads through strength testing on soil
samples, and the application of a percentage of these maximum loads on these
soil samples. Creep data such as
axial and volumetric strains will be measured and studied over a long period of
time.
Strength Testing
Consolidated undrained triaxial tests were conducted on Manoa clay samples to determine its shear strength parameters. Samples were normally consolidated to varying effective stresses. The results of these tests are presented in Figure 1.
Results of the tests can be represented with stress paths and effective stress failure envelopes. These methods are presented in Figures 2 and 3. Using either method, predictions can be made on the strength behavior of the soil under different effective stresses in drained or undrained conditions.
Creep samples were normally consolidated and subjected to constant loads under drained conditions in triaxial cells. Using the strength data, the maximum deviatoric stress under drained conditions at a given effective stress can be predicted. Creep samples are then subjected to a percentage of that maximum load. The actual load is applied very slowly at a constant rate to allow for pore pressure dissipation. Creep tests conducted so far are presented in Figure 4.
Landslide Modeling
Alani-Paty Landslide
Like other slow-moving landslides in Honolulu, the Alani-Paty landslide in the eastern Manoa valley takes place in the vicinity of the gently sloping valley floor. Figure 6 shows the approximate boundaries of this landslide. Stability in the area has caused costly damages to nearby properties and public facilities. Slope stability analysis was performed along cross-section AA'. The results are also useful to assess the effectiveness of remedial work, such as the new drainage system and structural stabilization measures. Given the fine-grained nature of the soils, creep deformations are also of concern. Creep predictions around location X are presented below.
The Alani-Paty landslide has movement strongly related to rainfall. The landslide material consists of coarse clasts (38% by volume) and sand embedded in a fine matrix of highly plastic silt and clay. Most of the landslide material is fully saturated. Shown in Figure 7 are the basic sets of soil parameters used for the stability analysis.
Figure 6: Approximate boundaries of the Alani-Paty landslide
Figure 7a: Variation of residual effective cohesion with depth
Figure 7b: Variation of residual effective frictional angle with depth
A comprehensive slope stability analysis has been performed
using the computer software package SLOPE/W. Slip surfaces within ±5 ft of the existing one
are used in the analysis. Figure 8
shows the geometry and slip surfaces adopted, and Figure 9 shows the resulting
safety factors. For the worst
groundwater scenario (i.e., groundwater table at the surface), the safety
factor for the slope using the Morgenstern-Price method is 0.76. As the groundwater table is lowered, the
slope becomes safer. In order for
the safety factor to be greater than 1, the groundwater table should be located
at least 11 ft below the ground surface.
Figure 8: Geometry and slip surfaces used in slope stability analysis
(along cross-section AA’)
Figure 9: Safety factors under different groundwater conditions
Like many slopes on the hillsides on the island of Oahu, the area is also undergoing continuous creep deformation. Numerical creep computations were performed using the finite element software package PLAXIS. With the water table at the ground surface, the slope surface is found to be moving downslope at a decreasing rate. The elapsed time from the onset of creep affects the resulting downslope displacement predictions. However, it is difficult to choose a starting point for the analysis. Shown in Figures 10 and 11 are surface creep predictions for the Alani-Paty landslide area near location X.
Figure 11: Predicted downslope creep displacement at location X