中文摘要：

提出一种基于蒙特卡洛模拟的岩质边坡可靠度设计方法,建立基于蒙特卡洛模拟的岩质边坡可靠度设计实用框架,将设计过程分成4个独立的子模块,整个设计过程清晰明了,便于工程师理解和掌握,在设计中允许工程师根据工程经验和实际工况灵活地调整目标失效概率和不确定性参数的概率模型。通过香港秀茂坪岩质天然边坡说明所提出的方法和设计框架,结果表明：（1）根据提出的方法通过执行一次MCS得到的设计空间内所有方案的失效概率与对每个设计方案重复执行MCS得到的失效概率十分一致,验证了所提出的方法的有效性;（2）目标失效概率以及不确定性参数的变异性和相关性显著影响边坡的最终设计方案,这种影响可以通过提出的方法合理地考虑到岩质边坡可靠度设计中;（3）在工程实际中,高估不确定性参数的变异性和忽略不确定性参数间的负相关性,均导致高估边坡稳定安全系数的标准差,进而低估边坡可靠度指标,导致相对保守的设计,偏于安全;然而,忽略不确定性参数间的正相关性则可能导致相反的结果,设计偏于不安全。

英文摘要：

This paper develops a reliability-based design（RBD） approach for rock slopes based on Monte Carlo simulation（MCS）,and proposed a practical design framework that is comprised of four design modules. Under the proposed design framework,the design process of MCS-based RBD of rock slopes is transparent and accessible to geotechnical practitioners so that they can make assumptions deemed appropriate for a specific site and have flexibility to adjust the target failure probability and uncertainty model. The proposed approach and design framework are illustrated using the Sau Mau Ping rock slope in Hong Kong. Results show that：（1） The failure probabilities of possible designs obtained from a single run of MCS in the proposed approach agree well with those obtained using repeated runs of MCS for different designs,which validates the proposed approach.（2） The target failure probability and uncertainty modeling may affect the final designs significantly. Such effects can be taken into account in RBD of rock slopes through the proposed approach in a rational manner.（3） In design practice,overestimation of the variability of uncertain parameters or ignoring the negative correlation between them might lead to overestimation of standard deviation of safety factor and,hence,underestimation of reliability index,which subsequently results in conservative designs. On the other hand,ignoring the positive correlation between uncertain parameters might lead to overestimation of reliability index and unsafe designs.