Comparision of stress variables performance in predicting the shear strength of unsaturated soils

Shear strength is one of the most important features in mechanical behavior of soils. The shear strength of unsaturated soils is still a controversial discussion between the researchers in this field. The methods of determining unsaturated shear strength are classified into two major categories; one of them employs two independent stress variables namely matric suction and net stress and saturated and unsaturated strength parameters are considered to be independent. In other words, as soon as the pore water pressure becomes negative, the saturated effective friction angle and cohesion become invalid. This approach became dominant especially since the validity of effective stress in unsaturated soils was questioned because it was not clear how to describe the collapse phenomenon through effective stress concept. In late 90’s some researchers referred back to effective stress concept and some ambiguity in explaining collapse was resolved. In this approach, effective stress is the main stress variable. Net stress and suction are combined into effective stress. The saturated and unsaturated shear strength parameters are assumed are not assumed to be independent from each other and there is a smooth transition between saturated and unsaturated soil modeling. In this research these two approaches are compared by means of unsaturated direct shear experiments and some relevant experimental data from literature. The advantages and shortcomings of the mentioned methods are analyzed. In the direct shear experiments, a wide range of soil suction was applied to the samples. Therefore it is possible to compare the effective stress and independent stress approaches in a wide range of suctions. The suctions of samples were measured by filter paper method. By plotting the failure envelopes in two approaches, the advantage of effective stress approach over the approach of independent stress variables is obvious. This advantage is especially drastic at higher suctions. The experimental data from literature similarly revealed this result. Thus it can be stated that effective stress approach is simpler and less time consuming since the failure envelope is a unique line for all suctions and strength parameters of a soil at saturated and unsaturated states are identical. On the contrary of independent stress variable approach, it is not required to measure strength parameters at various suctions. In the other words, if the effective stress is properly estimated, the unsaturated shear strength can be predicted straightforwardly. Effective stress parameter is the key factor for appropriate evaluation of effective stress in unsaturated soils. One of the highly cited proposed equations for effective stress parameter is verified by experimental data. The values of predicted effective stress parameter and the values measured from experiment are plotted versus suction. There is a good agreement between the effective stress parameters calculated from the equation and those measured from experimental data. Therefore it can be concluded that the empirical equation can accurately predict the effective stress parameter. It is worth mentioning that by normalizing the suction through dividing it into air entry suction, the effective stress parameter versus normalized suction becomes a unique line regardless of soil type. Thus the effect of soil type and its structure is normalized by means of using suction ratio.