AN ALTERNATIVE CENTRIFUGE METHOD TO OBTAIN THE SOIL -WATER CHARACTERISTIC CURVES FOR FINE -GRAINED SOILS

This thesis discusses the theory and a test procedure for use of a small-scale medical centrifuge to obtain the soil-water characteristic curve for statically compacted fine-grained soils. A centrifuge soil specimen holder was previously designed at the University of Saskatchewan for testing Beaver creek sand, Processed silt and Regina clay. This previous soil specimen holder was modified and the experimental procedure was altered to obtain the soil-water characteristic curve for fine-grained soils. The soils tested during this study were Indian Head till, processed silt and Regina clay. Statically compacted specimens of the above-mentioned soils were used for obtaining the centrifuge soil-water characteristic curve as compared to slurried specimens that were used in an earlier study done on Beaver creek sand, processed silt and Regina clay using the same centrifuge. The soil-water characteristic curve defines the relationship between the amount of water in the soil and soil suction. It provides important information concerning the water holding capacity of a particular soil at different suction values. Current testing techniques such as Pressure Plate apparatus and Tempe cells have been generally used to obtain the soil-water characteristic curve. These apparatuses are quite reliable but require considerable time. During this study, statically compacted soil specimens were tested by using both the Tempe cells and the small-scale medical centrifuge.The time required to produce the soil-water characteristic curve (within the range of 0 to 500 kPa) by using the centrifuge was much shorter than the time required by the conventional Tempe cell method. The centrifuge soil-water characteristic curve was obtained in 24 hours for three types of Indian head till specimens (13%, 16.3% and 19.2% initial water content), in 12 hours for the processed silt specimens (22% initial water content) and in 48 hours for the Regina clay specimens (38% initial water content). In comparison, the Tempe cell tests done on the similar three types of Indian head till specimens (13%, 16.3% and 19.2% initial water content) took 4, 5 and 6 weeks, respectively,to produce the soil-water characteristic curve within the same suction range.(i.e., 0 t0500 kPa). The Processed silt specimens took 2 weeks to produce a soil-water characteristic curve using the Tempe cell method and the Regina clay specimens took 16 weeks to produce the soil-water characteristic curve using the Pressure plate method. The time stated above for the centrifuge tests corresponds to the time the soil specimens were rotated in the centrifuge with speeds of 300 rpm, 500 rpm, 1000 rpm, 1500 rpm, 2000 rpm and 2500 rpm in one single test run. The proposed centrifuge method can be used to obtain multiple water content versus suction data points (twelve in this study)on the soil-water characteristic curve in one singe test run, in comparison with the conventional pressure plate and Tempe cell methods. Thus, the centrifuge method can be used with considerable time saving for obtaining the soil-water characteristic curve for statically compacted fine-grained soils. The statically compacted soil specimens were used to reduce the amount of compression the specimens would experience during centrifugation due to the increased gravity field. The results obtained by both the Tempe cell tests and the centrifuge tests on the similar soil, match closely. Therefore, the proposed centrifuge method can be used to obtain the soil-water characteristic curve for statically compacted fine-grained soils