Effect of triaxial specimen size on engineering design and analysis

Abstract

Triaxial tests are often used to determine the behavior and strength characteristics of soils. Specimen size can have a significant impact on the measured shear strength. Accordingly, the selected parameters affect the related geotechnical engineering analysis and design. We tested three different specimen sizes of loose Ottawa sand in triaxial compression tests. The measured shear strength and friction angle are used to explain some of the observed scale effects in engineering design and analysis. Critical state parameter and shear strength from the laboratory tests are employed to assess the static and seismic slope stability of an embankment dam, to calibrate a critical state soil constitutive model, to study the soil behavior under shallow foundations, and to evaluate liquefaction triggering and failure of retaining structures. We show that all of these analyses can be significantly affected by the choice of the specimen size used to determine shear strength parameters. While using small size samples for determining shear strength parameters might result in un-conservative design, a large sample size is consequently a more accurate representation of soil strength conditions and field deformations.