突發及漸進式觸發顆粒流崩塌動力之實驗

Abstract

本論文的動機源於對於布唐布納斯溪流域和荖濃溪地區的崩塌現象的觀察，其中存在著兩種明顯不同的崩塌類型：山崖退縮崩塌和侵蝕與堆積的崩塌。儘管現地數據能夠提供有關崩塌發生前後的資訊，但對於崩塌過程卻無法得知。因此，本論文旨在通過理想化的實驗來模擬和研究這些崩塌過程。為了研究這兩種崩塌類型的特點，我們設計並進行了傾斜通道實驗和旋轉滾筒實驗。在這兩種實驗中，我們使用石蹓砂作為崩塌材料，以模擬真實崩塌的行為。在傾斜渠道實驗中，崩塌是通過移除坡道下游端的擋板來觸發的，進而引起回退侵蝕。在旋轉滾筒實驗中，崩塌是通過逐漸增加的坡度觸發的。這種累積持續到達到臨界點，最終導致崩塌的發生。此外，這些實驗還考慮了崩塌尺度的變化，通過調整參數，如傾斜通道實驗中的出口高度和旋轉滾筒實驗中的崩塌長度。這使我們能夠研究不同因素對於崩塌過程和結果的影響。在傾斜通道實驗中的實驗結果顯示，下游端的擋板大小與斜坡後退速率之間存在明顯相關性。具體而言，越大的擋板移除，斜坡後退速率越快，並且產生的波形變得不那麼明顯。另一方面，在旋轉滾筒實驗中，我們觀察到崩塌的最大表面速度發生在崩塌事件的後半段。這些實驗結果能對未來發展顆粒流理論奠下基礎。

This thesis aims to replicate and study landslide processes through idealized laboratory experiments. The study is motivated by observations of landslides in the Pu-Tan-Pu-Nas watershed and Laonong River area, where two distinct types of landslides have been identified: scarp retreat subparallel landslides and landslides involving material entrainment and detrainment. While field data provide valuable information on post-landslide conditions, they lack insights into landslide processes. To investigate the characteristics of these two types of landslides, we designed and conducted inclined channel experiments and rotating drum experiments. In both types of experiments, garnet sand is used as a surrogate material to simulate the behavior of real landslides. In inclined channel experiments, the avalanche is triggered by the removal of baffle plates from the downstream end of the slope, causing retrogressive erosion. In rotating drum experiments, the avalanche is triggered by the gradually increasing slope. This accumulation continues until a critical point is reached, and eventually causes an avalanche. Furthermore, the experiments consider variations in landslide scales by adjusting parameters such as outlet height in inclined channel experiments and avalanche length in rotating drum experiments. This allows us to examine how different factors influence landslide processes and outcomes. The experimental results from the inclined channel experiments reveal a clear influence of the baffle size. Specifically, larger baffle plates lead to a faster scarp retreat, and the resulting wave shape becomes less distinct. On the other hand, in the rotating drum experiments, we observed that the maximum surface velocity of the avalanche occurs during the latter half of the avalanche event. These experimental findings lay the foundation for future developments in granular flow theory.