土体饱和渗透系数表现为天然的空间强变异性,其均值通常沿土层深度方向呈递减趋势,但目前鲜有文献考虑这一特性。为此,以FLAC为平台,利用fish语言基于局部平均法建立了表征土体饱和渗透系数均值随深度递减的一维非平稳随机场模型,采用FLAC两相流模块模拟雨水的入渗过程,以蒙特卡罗法为框架,结合考虑正孔隙水压力的广义有效应力无限边坡稳定模型,探讨了饱和渗透系数空间变异性在不同降雨时刻下对边坡最危险滑裂面分布规律以及相应的边坡破坏概率的影响。结果表明:忽略饱和渗透系数的空间变异性的确定分析方法将不能真实反映边坡的安全性;随着降雨持时的增加,边坡最危险滑裂面发生在坡底基岩处的概率逐渐降低;随着饱和渗透系数竖向相关距离的增加,边坡最危险滑裂面发生在坡底的概率逐渐增加,而相应的边坡破坏概率却随竖向相关距离的增加而逐渐降低。
The great natural spatial variability whose expectation value decreasing with the soil depth can be observed for saturated hydraulic conductivity. Therefore, the local average method is adopted to simulate the corresponding one- dimension non-stationary random field of spatial varied hydraulic conductivity using Fish code in FLAC. The process of rainfall infiltration is calculated using the module of two-phase flow. Within the Monte-Carlo framework, the influences of variability of saturated hydraulic conductivity on the distribution of the critical slip surface and on the corresponding failure probability at different time-moment of rainfall were studied, based on the infinite slope stability model with the effects of positive pore water pressure considered. The results show that without considering the variability of saturated hydraulic conductivity, the deterministic analysis method fails to represent the realistic slope stability. With the increase of rainfall duration, the probability of critical slip surface occurring at the bedrock of slope base may decrease ; with the increase of vertical correlation distance for saturated hydraulic conductivity, the probability of critical slip surface occurring at the slope base may increase, but the corresponding failure probability of slope may decrease with the increase of vertical correlation distance.