Microstructure-based estimation of unsaturated shear strength from simple evaporation test

Abstract

The effect of microstructure on the hydromechanical behavior of unsaturated soils is an issue that is frequently addressed in literature. A simple procedure is proposed in this study to characterize the soil microstructure through evaporation tests. Time-series gravimetric water content data together with the drying rate curve of an evaporated soil specimen are used to reveal soil’s microstructural properties. The experimental output is used for calibrating the parameter for shear strength model proposed by Alonso et al. (2010). A test program consisting of constant suction and constant water content triaxial tests on various types of soils from non-plastic to highly plastic has been carried out. The shear strength of soils determined from the experiments is predicted with the shear strength model that is calibrated through evaporation tests on the same soils. The model predictions have matched the experimental results satisfactorily. The shear strength of the unsaturated soils is predicted from a quite simple experimental procedure.

Appendix

The labels of data series are identified as the first letter denotes the soil type such that “M” for Mersin silt, “E” for Elazığ silt, and “A” for Ankara clay. The second and third letters are used for identification of test type, “CD” for constant suction triaxial shear, and “CW” for constant water content shear. In the next letters, the net stress (n) and suction (s) are given with the numerical values (in kPa); in case of constant water content tests, the suction value corresponds to the initial value of suction at the beginning of the shear. The test program on Mersin silt consists of 9 constant suction and 5 constant water content triaxial tests. In Fig. 12, a secondary axis on the right-hand side was added for constant suction tests on 400 kPa net stress with different scale in order to better illustrate the deviatoric stress versus axial strain behavior of remaining tests which are accumulated within the narrower range between 100 and 500 kPa. The variation of degree of saturation under constant suction and constant water content shear is given in Figs. 13 and 14. Figure 14 shows CW two experiments on Mersin silt specimens having almost the same initial suction and confining stress but different initial degree of saturation. The maximum deviatoric stress reached by these two specimens is different in Fig. 12; MCWn25s272 specimen that has lower initial degree of saturation reaches higher deviatoric stress than MCWn27s270 specimen. In Fig. 14, the suction reduction in MCWn27s270 having lower initial degree of saturation is consistently less than that in MCWn25s272. As it was discussed in Gallipoli and Bruno (2017) stabilizing effect of water meniscuses increase with number of meniscuses per unit volume of solids and the intensity of the pull of the meniscuses. The intensity of the pulling by meniscuses is quite similar for MCWn25s272 and MCWn27s270 specimens having same initial suction. MCWn27s270 specimen has higher initial degree of saturation and so the higher number of meniscuses. Theoretically, higher deviatoric stresses would be reached in drained or constant suction shear of MCWn27s270 specimen. But, to shear specimens under constant water content, the condition has provided opposite outcome since the higher the initial degree of saturation, the higher the suction reduction and thus the less the pulling effect of the meniscuses. The variation of void ratio during triaxial shearing of Mersin silt is shown in (Fig. 15).

Fig. 12

Deviatoric stress versus axial strain plot of unsaturated triaxial tests on Mersin silt specimens (discrete lines are represented on the axis at the right-hand side concerning better illustration)

Fig. 13

Variation of degree of saturation in constant suction shearing of Mersin silt

Fig. 14

Variation of degree of saturation (continuous lines represented on the left axis) and suction (dashed lines represented on the right axis) during constant water content testing of Mersin silt. For the test MCWn98s50, external data logger used for pore water pressure measurement remained disabled for a while at level of 30 kPa suction; final suction of 24 kPa is merged to previous measurement

Fig.15

Variation of void ratio during triaxial shearing of Mersin silt

The test program for Elazığ silt consists of 4 constant suction and 4 constant water content triaxial tests. The experimental results of constant suction and constant water content shearing of Elazığ silt specimens were given in following Figs. 16, 17, and 18

Fig. 16

Deviatoric stress versus axial strain behavior of Elazığ silt

Fig. 17

Variation of degree of saturation (on the left axis with continuous lines) and suction for CW tests (on the right axis with dashed lines) during triaxial shearing of Elazığ silt

Fig. 18

Variation of void ratio for Elazığ silt specimens

The results of 4 constant water content triaxial tests on Ankara clay specimens were illustrated in Figs. 19, 20, and 21.

Fig. 19

Deviatoric behavior of constant water content triaxial testing of Ankara clay

Fig. 20

Variation of degree of saturation (on the left axis) and suction (on the right axis with dashed lines) of Ankara clay during constant water content shear

Fig. 21

Variation of void ratio of Ankara clay during constant water content shear