地震液化的突变模型研究
A REVISED CUSP CATASTROPHE MODEL FOR PREDICTING LIQUEFACTION POTENTIAL OF SOILS INDUCED BY EARTHQUAKE
-
摘要: 能量和标准贯入击数是用于液化评价的两个综合性指标。以经过持时修正的能量函数和经过粘粒及上覆压力修正的标贯击数为控制变量,以振动孔隙水压力增长为状态变量, 根据液化的物理意义、室内试验所得到的孔压发展曲线的特点和国内外实测液化资料,建立了基于修正Zeeman 突变方程的孔压发展模型。与目前广泛使用的传统尖点突变模型相比,修正模型增加了一项和三个系数,这样做既保留了原方程中状态变量与其中一个控制变量成线性关系这一优良性质,又能够适应于准确模拟孔压与液化发展。在此基础上,利用液化资料,基于统计学中控制“弃真”概率的方法,运用Matlab 逐步逼近确定了所有的六个待定未知量, 确定了分叉曲线和新旧坐标之间的转换关系。检验结果表明了所建议模型的合理性。Abstract: The shaking-induced vibration energy and below count of standard penetration tests are two key factors for evaluating the liquefaction potential of sand. A revised model based on Zeeman equation of cusp catastrophe, which can display the development of liquefaction, is established. In the proposed model, the ratio of shaking-induced pore water pressure with effective confining pressure is taken as the state variable while the earthquake-induced energy function modified by earthquake duration and the below count of SPT corrected by the clay content and overburden confining pressure are taken as two individual controlling variables. In the study, triggering condition of liquefaction based on pore water pressure, development pattern of pore water pressure induced by cyclic loading under undrained condition achieved by experimental tests and seismically-incuced damage records of liquefaction observed in previous earthquakes are taken into account. Compared with the conventional cusp model widely employed in engineering practice, an additional term and three non-unity coefficients are included in the revised model. Therefore the fundamental feature of Zeeman equation and the linear correlation between two independent variables are fully contained in the proposed model and furthermore the model can adequately simulate the liquefaction development. Based on the proposed model, the statistical method based on the validation hypothesis for controlling first type of error is used to define the related parameters, bifurcated curve and transformation relationship between the two coordinate systems. The proposed model is verified by the test results. It is shown that the proposed model is reasonable and can be used to predict liquidation potential of sites in practice.