數值模擬結合水槽模型試驗於土石流堆積之研究

A Study on Numerical Simulation and Flume Experiments of the Debris Flow Deposition

葉姿君(Tzu-Chun Yeh)

指導教授 ： 林基源

朝陽科技大學/理工學院/營建工程系/碩士(2009年)

摘要

摘要
目前土石流的模擬及研究上大多數是以室內水槽試驗進行對現地土石流情況的模擬，或是以O’Brien發表之FLO-2D二維洪水及土石流套裝軟體針對現地之雨量資料及數值地形等資料模擬土石流之流動狀況。因此，本次研究利用FLO-2D配合室內水槽網格、實驗數據與相關參數進行數值分析，模擬土石流之流動情形，藉由數值程式模擬室內水槽之無攔阻試驗將模擬所得之堆積深度、堆積形狀和流動速度的探討來加強室內試驗與現地的關聯性，提供室內水槽試驗驗證的一種依據。
本研究主要分成兩部份，第一部份為數值高程模型的建立；第二部份為程式模擬結果之分析。
綜合各項結果可知坡度越大流速越快，但對於堆積的長度、寬度及深度並無明顯之影響，但在堆積影響範圍的準確度則隨著坡度變大而下降；說明在室內水槽試驗中坡度的大小會影響堆積的形狀及範圍，而在數值模擬中這樣的情況相對而言並不明顯。從數值分析之結果可發現體積濃度（Cv）的改變會對流速及體積有顯著之影響，Cv值越大體積隨之增加，而流速會隨之變慢，但對堆積範圍的影響則不明顯；這樣的情況顯示Cv值會影響流速及堆積的深度，但並不影響堆積長度及寬度的發展。
在準確度因子Ia（影響範圍之相似度）的分析上可知縮放比例係數50倍＞25倍＞10倍；在流速分析的比對上未經尺度分析前皆能符合現地土石流之流速範圍，即使尺度分析後大部份之結果亦都符合實驗是土石流之流速。從流速及準確度分析結果中可知在縮放比例係數10倍時略顯粗糙，精度與另外兩種倍數比較下略顯不足。

關鍵字

室內水槽試驗 ； FLO-2D ； 土石流

並列摘要

Abstract
Currently, most of debris-flow simulations and studies are those simulations of on-site debris flow situations conducted by the indoor tank test or those debris flow movements simulated by FLO-2D software, issued by O’Brien, in the light of on-site rainfall and digital terrain data. Therefore, this research uses FLO-2D with indoor tank mesh, experimental data and related parameters to make a digital analysis and to simulate the debris-flow movement. With the simulation of non-retaining test in an indoor tank by the digital program, we investigate its deposition depth, deposition form and velocity to reinforce the correlation between indoor and on-site tests to provide a verification basis for the indoor tank test.
This research can be divided into two parts: the first is the establishment of digital elevation model (DEM) and the second is the analysis of program-simulated results.
In conclusion, the result shows that the bigger the gradient is, the higher the velocity is but there is not obvious effect on the length, the width and the depth of depositions. However, the accuracy of deposition ranges will be declined as the gradient is increased, showing the degree of gradient will influence deposition forms and ranges in the indoor tank test. But it is not evident for such situation in the digital simulation. From the result of digital analysis, it is found that the change of volume concentration reduction rate (Cv) will have an effect on the velocity and the volume. The bigger the Cv is the increasing the volume is and the lower the velocity is. But there is not obvious effect on the deposition range. The above situation shows that the value of Cv will affect the velocity and the deposition depth but will not affect the development of deposition length and width.
From the analysis of accuracy index, Ia (the similarity of impact range), it is known that the scaling factor is: 50 times＞25 times＞10 times. For the comparison of velocity analysis, it tallies with the velocity range of on-site debris flows before the dimensional analysis. It also tallies with the debris-flow velocity in the laboratory even after the dimensional analysis. From the result of velocity and accuracy analyses, it is found that it is slightly rough for the 10-time scaling factor and its accuracy is a little improper in comparison with the other two multiples.

並列關鍵字

Indoor Tank Test ； FLO-2D ； Debris Flow

參考文獻

3 行政院農業委員會，「水土保持手冊」，1992。

17 陳文福、李毅宏和吳輝龍，「結合地文與降雨條件以判定土石流發生之研究－以陳有蘭溪集水區為例」，台灣地理資訊學刊，第二期：27-44 ，2005。

19 連惠邦、程向榮、李木青，「矩形防砂壩流量公式之研究」，中華水土保持學報第三十卷第二期，第127-135頁，1999。

22 彭繼賢，「應用FLO-2D於臺灣中部地區土石流流況分析之研究」，碩士論文，臺灣大學土木工程學研究所，2006。

26 趙啟宏，「土石流之數值模擬及流變參數特性之探討」，碩士論文，台灣國立大學土木工程研究所，2004。