Response spectra of slender columns as obtained with 2D-DDA and geophysical site response tests

The capability of the numerical DDA method to perform site response analysis is tested. We compare the resonance frequency obtained for a multi-drum column when modeling it with DDA and testing it in the field. When the numerical control parameters are properly selected we obtain a good agreement between DDA and the geophysical site response survey in the field. We find that the choice of the contact spring stiffness, or the numerical penalty, is directly related to the obtained resonance frequency mode in DDA. The best agreement with the physical test is obtained with a relatively soft contact spring stiffness of k = (1/20) E_o*L_o, where E _o is the Young's modulus of the blocks and L_o is the average block diameter. This optimal k value falls within the range of acceptable k values obtained in preliminary calibration tests performed independently of the field data. We find that the obtained resonance frequency is independent of the time step interval selected. Furthermore, DDA returns only a single resonance mode whereas the geophysical test results indicate two modes. This discrepancy is explained by the fixed base used in the DDA model which inhibits soil structure interactions.