Failure Analysis of Jointed Rock Slopes Based on a Modified Smooth-Joint Model

岩石邊坡常含有多組節理面，其為影響岩石邊坡穩定性的關鍵。目前分析節理岩坡穩定性的主要方式，多視其為連續體，並以極限平衡分析或連續體數值分析等方法進行探討，然而，對於分析節理岩坡中岩塊滑動破壞及塊體大量位移問題，節理面的力學行為具有極大的影響力，因此，透過能考慮分離塊體力學行為的離散元素法進行節理岩坡之穩定性分析，將可提供一更精確之預測。 本研究應用PFC3D軟體作為離散元素分析之工具，然而，因PFC在模擬節理面時過度簡化，無法合理反映具粗糙度節理面之力學行為，故本研究透過本團隊提出的修正平滑節理模式(Modified Smooth-Joint Model, 簡稱MSJ模式)模擬節理面力學特性。MSJ模式以Barton之剪力強度準則為基礎，由各項節理參數決定其力學特性。研究將先以數值直剪試驗驗證MSJ模式之準確性，接著建置簡易岩坡模型，探討其於極限平衡分析與節理參數之關係，然後再探討節理連續性之影響，並討論MSJ模式應用於岩坡破壞分析之可行性。 在驗證MSJ模式部分，研究首先由單軸壓縮試驗結果檢驗完整岩石微觀參數的正確性；接續以直剪試驗結果驗證節理面參數之合理性；最後檢驗直剪模擬結果與Barton之剪力強度模式是否相符，並與未修正之SJ模式互相對照。試驗結果經比較後，發現MSJ模式能合理表現出Barton模式中尖峰剪力強度隨正向應力變化之趨勢，而SJ模式在較粗糙的節理面則無法呈現出差異。 簡易岩坡分析部分，研究將以單一節理之梯形邊坡進行。首先依極限平衡理論，改變節理摩擦角參數設置，分析節理傾角與摩擦角關係；再來以改良的MSJ模式，探討節理面上摩擦角空間分佈差異，藉此瞭解節理面上正向應力作用的影響；最後將連續的節理面改為具岩橋之不連續節理，透過岩坡之變形及破壞分析，依序探討節理連續性、完整岩石強度與岩橋型式對岩坡穩定性造成的影響。 透過岩坡的各項數值模擬結果發現，在岩坡中，即使節理摩擦角小於傾角，坡體也未必發生滑動，此異於以往無限邊坡之極限平衡分析；節理面摩擦角會隨上部塊體重量分佈而改變，此特性反映出具粗糙度節理面力學行為受正向應力之影響；若節理面不連續，則節理破裂發展與岩橋長度與型式有關，且岩橋破壞將導致上部塊體滑動。未來可進一步模擬實際岩坡，以利於工程分析與設計之評估。Rock slope often contains discontinuities such as joints, which is the key factor of rock slope stability. Conventionally, the limit equilibrium analysis and continuum numerical analysis are most used method for stability of jointed rock slope. However, the mechanical behavior of joint surface influence the sliding failure of rock block and massive transfer of block in jointed rock slope extremely. Therefore, the discrete element analysis which can consider the mechanical behavior of discrete block is an effective way to investigate the stability of jointed rock slope. In this study, PFC3D software is choose as the discrete element analysis tool, however, PFC is over-simplify in simulate joint surface behavior, thus it cannot react the mechanical behavior of roughness joint surface. Therefore, this study simulates the mechanical characteristic of joint surface through the improved method - Modified Smooth-Joint Model (MSJ model) which is developed by our research team. The MSJ model is based on Barton''s shear strength criterion, the mechanical behaviors are determined by various joint parameters. This study first verify out the accuracy of MSJ model by numerical direct shear tests, then we construct a simple rock slope model to discuss the relationships between limit equilibrium analysis and joint parameters, and then the influence of joint continuity is investigated, finally discuss about the application of the MSJ model on rock slope failure analysis. In the part of MSJ model verification, we first to examine the reliability of intact rock micro-parameters through the simulation data of uniaxial compress test; then the rationality of joint parameters are verified by the results of direct shear test; finally, the simulation results of direct shear test and Barton''s shear strength model are compared to ensure that they are equal, and the SJ model is treated as the control group. The comparison shows MSJ model can reasonably represent the trend of Barton''s model which shows friction angle is correlate with normal stress, while SJ model cannot point out this important characteristic in roughness joint. In the section of simple rock slope analysis, trapezoid slope with single joint is used in the simulation. The first step is changing the joint friction angle parameters to analyze the relationship between the joint friction angle and dip angle according to the limit equilibrium theory; second, the revised MSJ model is used to investigate the spatial distribution of friction angle to recognize the influence of normal stress acted on joint surface; finally, the continuous joints are alter to discontinuous joints to simulate the effect of rock bridge, and then sequentially investigate the influence of joint persistence, intact rock strength and rock bridge type to the rock slope stability through failure analysis of rock slope. According to the simulation results of simple rock slope, there are some conclusions and suggestions of the MSJ model. The simulation results show that even joint friction angle is lower to dip angle, the slope is not necessarily to slip in numerical rock slope. We suggest that three-dimensional limit equilibrium analysis can be adapted in future to correct the settings of numerical model; joint friction angles is vary with the distribution of upper block weight, this feature shows the mechanical behavior of roughness joint is influence by normal stress; if joint surface is discontinuous, the crack distribution are correlate to the length and type of rock bridges, and the failure of rock bridge will cause upper block to slide. We recommended that MSJ model can be used to simulate real rock slope in the future, and exploring the complex mechanism of jointed rock slope failure to evaluate the analysis and design of geo-engineering