Effects of specimen size and inertia on resonant column tests applied to sands

Abstract

The availability of a Stokoe type resonant column apparatus adapted to accommodate specimens of different size motivated a closer look into the details of the calibration procedure, with the aim to refine it so as to ensure that the derived shear moduli are independent of the specimen size. The starting point for the investigation is the experimental evidence reported by previous studies that the effective drive head inertia depends on the measured resonant frequency. In the paper, the theoretical background of the data reduction procedure is re-visited, and two calibration approaches are analyzed and applied for three distinct specimen sizes: i) the common technique using added masses, and ii) one without added masses relying on the system response with presumed calibration bar dynamic properties. The geometry of the calibration bars is varied, and the effects of non-planar torsion are assessed. The first approach, with frequency and specimen size dependent drive head inertia, is subsequently implemented in test series on three sands with different grain size distribution curves. The specimen size is varied. The experimental results in terms of shear modulus and damping ratio at small and intermediate shear strains yield almost coinciding curves for all specimen sizes considered, thus confirming the adequacy of the data reduction procedure. Known features of the dynamic soil behavior are well reproduced. The agreement with results for similar sands reported in the literature is also very good. • Theoretical background of evaluation and data reduction procedure are revisited. • Calibration techniques with/without added masses are analyzed and assessed. • Necessity of drive head inertia dependent on frequency and specimen size is shown. • Range of eigenfrequency in calibration must comply to that anticipated in the soil tests. • Proper calibration yields modulus and damping largely independent of specimen size.