چكيده به لاتين
In this study, a centrifuge test was numerically modeled and analyzed using sand constitutive models developed in OpenSees software including SANISAND, PDMY, PDMY02, PDMY03, CycLiqCP and CycLiqCPSP, which have the capability to model liquefaction. This centrifuge test is based on a offshore wind turbine with gravity foundation under a seismic load with liquefaction in the Touyra sand. The response of the numerical model to seismic stimulation including the ratio of excess pore water pressure, acceleration history, lateral displacement, foundation settlement and free soil settlement was reported for each constitutive model. Estimation of these responses was discussed. In this comparison, all conditions including dimensions and type of mesh, number of elements, type of element, loading, damping, permeability coefficient, etc. are considered the same. After observing the results, it was found that constitutive models have many errors in estimating settlements and lateral displacement. The first step to investigate the cause of incorrect prediction of constitutive models in liquefaction conditions is to examine their parameters, which gives a good view of the effect of the parameters on the responses of the numerical model. Among the behavioral models used in this study, SANISAND (MD04) and Pressure Dependent Multi Yeildsurface (PDMY02) models have recently been considered by many researchers and have shown relatively good results; Therefore, in this study, the parameters of these two constitutive models have been studied in detail. The results of this parametric study determined the amount and type of effect of each parameter on the response
of the numerical model. These results also help to verify numerical models, for example, by knowing the important parameters and the effect on the pore water pressure, it is easier to correct and adapt it to the laboratory value. Finally, it was found that by changing the parameters, the values of the session do not change much and the reason for this is the low shear strain is predicted that no parameter has an effective role in increasing it
