Comparison of undrained shear strength by pressuremeter and other tests, and numerical assessment of the effect of finite probe length in pressuremeter tests

Abstract

There is a wide body of research reported in the literature which compare the undrained shear strength determined from various field and laboratory tests. The pressuremeter testing equipment and the related design techniques have been continuously refined. However, the undrained shear strength (S u) determined from the pressuremeter test (PMT) is generally higher than that obtained from other field or laboratory tests. In this study, numerical methods to consider the effects of length to diameter ratio (L/D) of the pressuremeter probe, and the procedural implications of soil disturbance and testing depth on the undrained shear strength determined from the PMT were investigated and possible correction factors depending on L/D ratio were established. In addition, in order to compare the variations in undrained shear strength of a clayey material with depth and testing method, a lightly overconsolidated and highly plastic clay, called “Eymir Lake clay” near Ankara (Turkey), was selected as the material of the study, and its values of undrained shear strength (S u), which were determined by PMT, field vane shear (FVT), conic penetration test (CPT), and laboratory tests, were compared. Based on the numerical analysis results, the correction factors depending on the L/D ratio of the conventional probes were suggested. The correction factors ranged from 0.83 to 0.96 for L/D ratios of 5.3 to 11, respectively, and they were determined to be independent of the rigidity values ranging from 25 to 200. It was also shown that overestimation of S u is independent of depth. In addition, due to very low permeability values of the Eymir Lake clay, it is concluded that for soils with coefficients of permeability lower than 10−10 m/s, partial drainage around the pressuremeter probe is unlikely. Based on the comparison among the results from laboratory and different field tests, the values of S u determined from the theoretical Palmer’s solution are higher than those from CPT, FVT, and CU tests. If the correction for L/D ratio is applied to the values of S u determined from the Palmer’s solution, S u from pressuremeter approaches S u obtained from FVT; however, there is still a slight overestimation in S u obtained from PMT. This overestimation is probably due to the differences in the mode of failure and the presence of a disturbed zone around the pressuremeter probe. However, S u determined from the PMT by empirical methods is close to those determined from CPT, FVT, and CU tests.