بهمن نیرومند

Damping ratio and normalized shear modulus are of the most important properties of soils. These parameters employed to solve the problems such as assessment of seismic response of the site, dynamic analyses, and soil-structure interaction. Correct and appropriate evaluation of these parameters is needed to achieve more accurate results and safer design. This study proposes several models to assess the dynamic parameters i.e. normalized shear modulus and damping ratio of sandy and clay soils by using generation expansion planning (GEP). It is also possible to model the unknown nonlinear hyperbolic parameter or HD that is Reference shear strain for sandy and clay soils by GEP. In order to model the unknown parameters of the two nonlinear one-dimensional model i.e. Ramberg–Osgood model and modified hyperbolic, we decided to use nonlinear regression in Matlab software on sandy and clay samples. This resulted into discovering the unknown parameters of the two models. Initially we used (GEP), just like hyperbolic model in order to obtain formulas to estimate the unknown parameters of the model as a function of the other soil parameters (e, OCR, PI). However the extreme scattering and considerable errors of outputs in (GEP), revealed that estimation of the model parameters does not have enough accuracy. Therefore, parameters estimation proceeded by two approaches: 1. Performing regression analysis on all data 2. Mean calculation of all regression data results. The data used in the modeling process are all collected from laboratory results of previous accredited researchers. The input parameters of these models introduced to software for sandy soils include shear strain percent, mean effective confining pressure, void ratio four parameters are considered as inputs for the models, including mean effective confining pressure, shear strain percent, plasticity index, and over consolidation ratio. These models were evaluated by statistical parameters same MSE, RMSE and R2 in ord