Scaled Memory Model Parameters And Applied Research

Firstly, the scaled memory (SM) model is introduced for dynamic nonlinear analysis of soil structures. The SM translates the nonlinear variation of tangential modules during monotonic loading into a piecewise linear distribution geometrical relationship, which can easily be modified to simulate hysteretic characteristics of soil behavior during cyclic loadings, it also can simulate the dynamic stress-strain response of arbitrary cyclic loadings.In order to determine the parameters of the SM model, the analysis program of nonlinear Levenberg－Marquardt least squares method is developed. The parameters of the SM model , such as ,,,,,, is defined by fitting test results. To compare the dynamic stress-strain relationships, modules attenuation and damp ratio curve between the tests and the calculated results by SM model, it is shown that the SM model is capable and practicable. Because the hysteretic loops are varied as the axial dynamic strain increased, so the SM model parameters will be changed during dynamic loading. The relationship between h0,r,s and axial dynamic strain is induced by the fitting test results. In addition, the parameter of binary scaled model (BSM) is discussed which mean the skeleton curve is different between tensile branch and compression branch. Lastly, a seismic response analysis program of embankment, which combines step-by-step scheme with unbalanced load transfer method, was developed. It has thought of the degradation of soils and can use the parameters of the SM model that has been calculated by the least squares program. Taking a embankment as example, a comparison between the incrementally-iterative numerical technique and the equivalent linearization approach was carried out .It is shown that distribution of dynamic response of dam is mainly alike, but the dynamic response of incrementally-iterative numerical technique using experimental parameter is greater than that of the equivalent linearization approach.