This investigation presents the results of a comparative experimental analysis on the tensile strength obtained from soil samples in both conditions undisturbed and disturbed. For undisturbed soil samples, the tensile strength was obtained from an own-designed testing apparatus of direct tension. In this condition, the geometry of the samples during the carving process was designed to generate a failure plane in a predisposed area of the sample. For disturbed soil samples, a mold was built based on the used by Kim and Hwang (Eng Geol 69:233–244, 2003). Five types of stiff sandy soils from the Aguascalientes Valley, Mexico, were used as testing materials. For the analysis, the tensile strength obtained from undisturbed soil samples was compared with that of disturbed compacted samples within the mold, and at a dry density similar to that of their natural conditions. Three testing processes in disturbed samples were separately performed to obtain and analyze tensile strength. These include: (a) varying the molding water content; (b) completely saturating the samples and allowing them to dry at natural climatic conditions, until the water content was similar to the measured conditions in undisturbed soil samples; and (c) observing the variation in tensile strength on one type of soil when the water content was increased from 9.2%, which is the water content that corresponds to undisturbed samples at the moment of their testing. The results obtained from disturbed soil samples after a drying process were numerically more similar to the results of undisturbed soil samples, in large part because the drying process allowed the carbonated cementing substances of the soil to maintain the cohesion of the particles.
Read full abstract