A Critical Review of Membrane and Filter Paper Correction Formulas for the Triaxial Testing of Soft Soils

Abstract

Abstract This experimental study addresses the effect of membrane and filter paper on measurements of deviator stress by performing triaxial compression tests on gelatin, silicon, and Speswhite clay specimens. The magnitude of this effect is assessed at large strains close to the critical state conditions for soft soils with an undrained shear strength ranging up to 40 kPa. This study shows that the corrections for the membrane and filter paper effect, given by ASTM and BS, can lead to a significant overestimation of the correction required when applied to typical values for deviator stress found in tests on soft soils. Particularly at large strains the correction associated with the presence of membrane and filter paper, according to ASTM and BS, can be more than half of the material’s undrained shear strength. This study also demonstrates that for soft specimens, the membrane correction is strain level–dependent and can be predicted by taking into account the buckling state of the membrane during shearing. In addition, the strength contributed by the membrane is found to be independent of cell pressure but greatly affected by the friction between the tested specimen and membrane. Furthermore, the correction for filter paper is shown to depend primarily on the wetting of the filter paper strips, and to a lesser extent on their orientation. The combination of wet and oblique filter paper strips is the most effective way of making accurate measurements of deviator stress. This results in a filter paper correction that can be as low as 1 kPa, irrespective of the strain level. Finally, alternative expressions for the membrane and filter paper corrections to be used in the triaxial testing of soft soils are presented herein.