中文摘要 由於台灣地質狀況不佳及環境營力旺盛,邊坡破壞已成為台灣主要的地質災害之一,土壤邊坡主要為弧型破壞模式。而極限平衡法為目前最常被使用的分析方法,其需假設土體為剛塑性材料及找尋許多嘗試破壞面中對應最小安全係數者即為最可能之破壞面。本文另以考慮變形性的FLAC程式並與基於極限平衡之STABL進行比較分析,以了解土壤變形性、張力強度、互層邊坡及坡頂加載之影響。 本文探討土壤邊坡之穩定性為主,探討參數為安全係數及破壞面位置。STABL 分析採用基於圓弧滑動面之Bishop簡化法,破壞面搜尋數目定為一萬個。FLAC程式之安全係數訂定是以剪力強度折減觀念所決定,再由程式所輸出之剪應變等值圖及位移向量圖來判斷邊坡可能之破壞面位置。所分析邊坡皆為低地下水位之莫爾庫倫材料,但考量均質土壤邊坡、互層邊坡及坡頂受局部均勻垂直荷重邊坡三類,分別進行穩定性分析,並比較 FLAC程式與STABL程式之分析結果,以及找出STABL程式於不同邊坡情況下可能之不適用範圍。坡高皆固定為 8m,但坡角變化於 0 至 90 度之間。 針對上述三類邊坡,由FLAC程式與STABL程式分析所得之安全係數及破壞面位置,可得以下結果:1) 以FLAC程式進行土壤邊坡穩定分析,發現張力強度、彈性模數及土體之變形量對分析的結果均有影響,此乃STABL程式所無法探討者;2) 當坡角增加時,安全係數隨之下降;3) 於互層土壤邊坡中,當下層為較軟弱之土壤時,在坡角較大的邊坡時,下層土壤易發生局部大變形之破壞,而使FLAC所求得之安全係數較STABL分析所得較低;4) 於坡頂受局部荷重的情況下,FLAC分析所求得之破壞面與STABL有明顯的不同,其主要在於FLAC程式考慮土體的變形量,而使得破壞面形態受荷重的影響較大。
ABSTRACT Because of its inferior geological conditions and active environmental agent, slope failures have been one of the most common geological disasters in Taiwan; and arc-shaped failure surfaces are usually observed in soil slopes. Among various slope stability analyses, the limit equilibrium methods (LEM) is most frequently used, in which a soil mass is assumed rigid-plastic and the most critical failure surface is the one having the least factor of safety (FS) among numerous trial failure surfaces. The objective of this thesis is to perform a comparative study of FLAC code and LEM-based STABL program, for exploring the influence of soil deformability, tensile strength, inter-bedded slope, and crest loading on slope stability. The focus of this thesis is on the stability of soil slopes, in terms of factor of safety (FS) and failure surface location. The simplified Bishop method for arc-failure surfaces and ten thousands of searching sliding surfaces are adopted in the STABL program. Determination of FS in FLAC is mainly based on the concept of reduction in shear strength, and the corresponding failure surface is fairly located from the shear strain contour and displacement vector plots. The slope examined is a Mohr-Coulomb medium having a low water table, and three slope conditions are taken in account: a homogenous mass, an inter-bedded mass, a crest-loaded slope. Both analysis results of FLAC and STABL for each condition was compared, in order to seek the probable situations in which STABL may not be suitable. The slope height is fixed to be 8m and the slope inclination varies from 0 to 90°. From the simulation results of FLAC and STABL for the aforementioned three slope conditions, the following conclusions can be drawn that: 1) both soil elastic modulus and tensile strength somewhat affect the slope stability determined by FLAC, which can not be captured by STABL; 2) as expected, FS decreases with increasing slope inclination; 3) in the inter-bedded slope case with a weaker lower layer, FLAC yields a lower FS than STABL, due to the formation of localized failure zone in the lower layer; 4) in the crest-loaded slope case, FLAC also yields a different failure surface from STABL, due to the large deformation near the loading base.