Small-strain stiffness of liquefiable sands: A comparison between bender elements and resonant-column tests

Abstract

Soil stiffness can be estimated by geophysical and dynamic testing methods. In the laboratory, the most common methods to measure the small-strain stiffness are the bender elements (BE) and resonant-column (RC) tests. This paper focuses on the comparison between the results of the small-strain stiffness of sands by BE and RC tests. For this purpose, an experimental program involving three liquefiable sandy soils (i.e., NB, TP-Lisbon and Toyoura sands) was carried out. Such program covered the measurement of the small-strain stiffness of these soils by BE in triaxial and RC apparatuses for different mean effective stress conditions. All tests were carried out on saturated soil specimens, which were remoulded using the air pluviation method for various relative densities. The experimental results were interpreted in terms of shear-wave velocity (Vs) and maximum shear modulus (Gmax) to derive the stress-dependency laws of these parameters. The experimental results revealed differences between Vs obtained from BE and RC tests, evidencing a clear effect of relative density on the shear-wave propagation. However, such a variation may be significantly reduced when a normalisation of Gmax in terms of a void ratio function F(e) is applied. As a result, this study demonstrated and validated the importance of accounting for the soil state conditions, for adequate compatibility of BE and RC tests in the estimate of the small-strain stiffness of liquefiable sands.