Strength loss of sensitive clay slopes is a common concern for soft soil sites. A useful method for studying this problem is through physical model tests. However, a problem that deters more widespread use of such tests is the difficulty of obtaining the large samples of natural sensitive soil that would be required for physical model tests. The work described in this paper was motivated by the need to develop practical procedures to produce artificial sensitive clay for use in model tests.
To reach this goal, some fundamental experiments were performed to explore the effects of cement content, water content, aging, and remolding after aging on the undrained shear strength. Two soils (San Francisco Bay Mud and Yolo Loam) with several different cement and water contents were studied. After mixing and aging, and measurement of the aged strength, the soils were again remolded to determine the aged-remolded undrained strength. The sensitivity of the aged clay could then be calculated by dividing the aged strength by the aged-remolded strength.
San Francisco Bay Mud (SFBM), a high plasticity clay, developed a sensitivity of approximately 3 to 3.5 after 7 days due to thixotropic hardening alone. Sensitivities of 4 to 8 were obtained by addition of a small percent (2 ? 5%) of cement to the soil. Yolo Loam (YL), a low plasticity lean clay rich in silt content, produced much higher sensitivities of 20 to 27 with a small addition of cement (2 ? 3%). A procedure for production of physical model test specimens with the desired strength and sensitivity is also outlined.