A simple procedure to calibrate a pore pressure energy-based model from in situ tests

The simultaneous generation, dissipation and redistribution of excess pore pressures within the layers of a soil deposit, due to a seismic event, can significantly modify the seismic response of the whole deposit. The reliable estimate of the excess pore pressure induced by shaking within the soil is important to predict the behaviour of the soil at a large scale, and consequently, earthquake effects on built environment. Recently, pore pressure energy-based models are developing. Despite several advantages, their calibration is generally complex. The paper aims to provide a simple calibration procedure of the pore pressure energy-based prediction model developed by Berrill and Davis (1985), in order to make easier and more common the use to practitioners. The energy-based model of Berrill and Davis (1985) has been calibrated in this study by means of a dataset of 46 undrained cyclic triaxial and simple shear tests carried out on different sandy soils. The best fitting procedure with the envelope of the experimental curves has been adopted. The experimental evidences show that the two parameters (a and b) on which the model depends, are linked and can be related to the results of CPT or SPT in situ tests. The paper introduces two relationships to compute the calibration parameters from the well-known equivalent cone tip resistance (qc1Ncs) or the corrected SPT blow count ((N1)60cs). The applicability of the proposed procedure at a large scale has been discussed interpreting the results of cyclic simple shear tests on undisturbed sandy specimens from an energetic perspective. The validity of the calibration procedure has been finally verified performing 1D non-linear site response analyses by means of DEEPSOIL code, reproducing two centrifuge tests and three case histories. The close matching between the simulations of the excess pore pressure time histories with the experimental data of the centrifuge tests, together with the simulated excess pore pressure profiles of three case histories compared with the results achieved by using another 1D non-linear code demonstrates the effectiveness of the simplified procedure to calibrate the pore pressure energy-based prediction model of Berrill and Davis (1985).