以物質點法探討光華崩塌地滑動面深度與運動行為

Abstract

邊坡災害對人們生命及財產造成極大危害，不同的環境條件加上各樣觸發因子的交互作用，常常複雜化邊坡的破壞機制，難以預測其後續的發展，使整治工程與防災政策逐年更改與替換，導致最後標準化的流程僅能大略的劃出警戒標準，難以將整治規劃具體呈現。為此完善的現地調查搭配長期有效的監測為重中之重，並且根據調查結果進行數值分析有效掌握其破壞機制的運作與預測未來滑動趨勢，以此對邊坡有具體的了解才能對症下藥。 本研究以光華崩塌地作為研究案例，此邊坡具有長期滑動行為且為潛在大規模崩塌的邊坡，透過分析其自2018年以來的監測數據、活動性與整治工程成效，以掌握其可能的破壞機制。然而由於邊坡變形過大，導致地中調查儀器損壞，加上下邊坡(分區2)太過陡峭，無法再進行鑽探取得分區二破碎岩層的分佈，進而無法準確掌握整體滑動面深度與滑動量體。為此本研究根據調查結果，應用物質點法(Material Point Method)建立數值模型探討光華崩塌地之深層滑動面位置，作為後續預測崩塌後土石運動行為的依據。透過假設四種不同深度之破碎岩層地質模型進行反算分析，將數值模擬結果比對地中及地表監測數據(前期傾斜管、多期DEM、航拍照及地表伸縮計)以驗證模型，取得下邊坡破碎岩層之合理深度與滑動面位置。 數值分析結果顯示下邊坡破碎岩層深度為60公尺之模型最符合監測資料。光華崩塌地的滑動面由兩道剪裂帶(或破壞面)構成，第一道剪裂帶位於在上邊坡(分區1)岩屑層底部，第二道剪裂帶貫穿上下邊坡位於深度40公尺之位置。上邊坡下滑程度除了會受到第一道剪裂帶滑動之影響外，也受到第二道剪裂帶影響而加劇下滑。此外邊坡的滑動行為可分為三階段，初期：岩屑層滑動，中期：破碎岩層滑動帶動岩屑塊體的雙層滑動行為，後期：分區二受束口地形限制停止滑動，漸進的使岩屑滑動停止。最後，考量到地下水抬昇為岩屑滑動一大因素，故本研究透過施加水壓力於土岩交界以模擬地下水位上升的效應，分析結果顯示當水位上升至-7 m岩屑層可能發生二次滑動。

Landslide has endangered people’s life and properties. Different environments and trigger factors often complicate the slope failure mechanism, making it difficult to predict the subsequent development and to apply effective mitigation measure. Therefore, an integral site investigation and long-term monitoring systems are of first importance. Besides, based on the field investigation results, the numerical analysis can be conducted to identify the slope failure mechanism and predict the post-failure behavior for better understanding the overall failure mechanism and influence distance of landslides. This research presents a case study of Guanghua landslide, which is a potential large-scale landslide. Information of the monitoring data, past landslide activity, and the improvement treatments were analyzed to understand the failure mechanism. However, due to the large deformation, subsurface monitor systems have become invalid. Hence, the numerical models using material point method were developed to investigate the depth of the failure surface and predict the post-failure behavior. Four geological models with different depth of the fractured rock were considered in the back analyses. The numerical results were compared with subsurface and ground surface monitoring data for model validation to obtain the reasonable depth of the failure surface of landslide. The numerical results indicated the model with 60 m depth of fractured rock conform to the monitoring data well, and two shear bands in Guanghua landslide developed; the deep one past through the upper (District 1) and lower (District 2) slope at a depth of 40 m. Besides, the subsidence in District 1 was influenced by not only the first shear band but also the depth of the fractured rock. The kinematic behavior of the landslide involved three stages during the entire landslide process. In the first stage, the colluvium mass slides due to the development of the first shear band. In the second stage, both colluvium and fractured rock were displaced by the development of two shear bands. In the third stage, District 2 stopped sliding due to the terrain restriction. Moreover, a parametric study was performed to evaluated the influence of the raising in ground water level (GWL) on the movement of the colluvium layer. The results revealed that second sliding in colluvium could happen above the depth of 7 m.