Abstract
Mechanical soil parameters are not constants and can be defined in various ways. Therefore, determination of their values for engineering practice is difficult. This problem is discussed based on results of piezoceramic element tests and triaxial tests (unconfined and confined) on loess specimens improved by compaction and sand admixture (20% by weight). The study indicated also the effectiveness of this simple method of loess stabilization. The influence of specimen size, draining conditions, stress and strain state, and different calculation methods on the evaluation of basic mechanical parameters were analyzed. The initial shear and Young's moduli, the degradation of secant moduli with strain, tangent moduli, and Poisson' ratio were determined. The results showed that the shear strength parameters are much less sensitive to the test variables than the stiffness parameters are. In triaxial tests, the strength criterion adopted, the sample size, and the drainage conditions influenced the measured value of cohesion, with a much smaller impact on the angle of internal friction. On the other hand, the adopted definition of the parameter and the range of strains had the greatest influence on the value of the stiffness modulus. Moreover, larger specimens were usually found to be stiffer.