Abstract
To verify the expressions developed by other researchers for separating the strength component resulting from volume change in shear (dilatancy), shear tests at constant volume as well as drained tests are necessary. In conventional undrained triaxial tests on granular materials, volume change of the soil skeleton occurs as a result of rubber membrane penetration. In the present investigation, conventional tests were modified to ensure a constant volume condition by compensating for the rubber membrane penetration.
Drained tests, conventional undrained tests, and modified undrained tests were conducted on a uniform medium sand, uniform coarse sand, and a gap-graded sand; the results are compared. Conventional undrained tests gave somewhat higher values of shear strength than those of modified undrained tests. The results of drained tests corrected for volume change compare well with those of modified undrained tests.