中文摘要：

构造不同加载次序的阶跃荷载进行应变控制循环单剪试验，并基于R-N模型与传统P-M模型进行土体材料损伤计算，对比分析计算值与试验值，验证R-N模型的有效性；基于R_N模型既能考虑荷载幅值和加载次序的特征，根据R．N非线性疲劳累积损伤理论推导并建立砂土震陷计算方法；利用GCTS测试系统对美国Filter净砂展开研究，进行404组剪应变控制动循环单剪试验，考虑荷载幅值、上覆荷载、砂土特性及地震动特性等影响，测定模型参数，并建立模型参数回归模型；同时，进行202组不同特性的真实地震荷载输入剪应变控制动循环单剪试验，测定不同j二况下的砂土试样在地震荷载作用下引起的竖向变形；最后，将基于本文方法预测的砂土震陷时程曲线和累积值与试验值对比分析，验证了该方法的正确性。

英文摘要：

The strain-controlled cyclic direct simple shear tests are conducted using step loading with drafted different loading sequences. Based on the R-N model and classic P-M model, the damage values of the soil materials are evaluated and compared with the test ones in order to verify the validity of R-N model. In view of the characters of R-N model which can reflect the influence of loading amplitude as well as loading sequence, the evaluating procedure for seismic compression in sands is derived according to R-N cumulative damage fatigue nonlinear theory. Using GCTS testing system, 404 groups of strain-controlled dynamic simple shear tests are carried out for the U.S. Filter clean sand. Considering the effects of loading amplitude, overburden pressure and relative density of sand, the parameters of the model are tested, and a regression model for parameters is established. Also 202 groups of strain-controlled dynamic simple shear tests using input real seismic loads with different properties are performed to determine the vertical deformations of sand samples due to seismic loading under different cases. Finally, both history curves and cumulative seismic compression values predicted by the proposed procedure are compared with those from tests for the purpose of verifying the validity.