塊體抽離沖蝕行為數值模擬

Abstract

台灣河川多具坡陡流急之特性，西部麓山帶河川又屬較為年輕之地層，岩床極易受水流沖蝕。河床面快速降低及地形持續變化，造成河道不穩定且威脅各種水工構造物之安全或使其喪失功能。 岩石河床之沖蝕行為與較為人知的沖積層河床沖蝕行為差異甚巨，岩石河床局部之沖蝕行為因地形或水流條件等差異，產生不同之沖蝕機制。在岩床岩體屬於較為破碎之節理岩體時，岩塊抽離機制（plucking）常成為沖蝕行為之主控機制，其沖蝕速率往往驚人，因此需要特別重視與深入探討。影響岩塊抽離機制之因子則受岩體弱面控制，弱面位態、間距、切割形成之塊體形狀、弱面特性等皆造成不同之沖蝕行為。 數值模擬方法俱備可任意控制變因、可大量進行分析，耗時較短之優點，又不受試體取樣與準備不易之困擾，本研究因而嘗試以數值模擬方法為工具，將數值模擬視為「虛擬沖蝕試驗」，代替實體試驗，針對河段尺度之富含節理岩體的虛擬岩床模型進行沖蝕模擬，以探討不同因子對岩塊抽離沖蝕機制之影響。沖蝕模擬分析相關結果如下： 1. 節理位態對沖蝕啟動剪應力之影響受順水流方向之塊體爬升面坡角控制，爬升面坡角提升造成啟動剪應力相對上升。沖蝕啟動後之沖蝕行為則並非由爬升面坡角控制，而由模型中最平緩之坡面主控。由此可見，沖蝕啟動難易程度與啟動後之行為的影響因子並不完全相同。 2. 節理間距愈大即代表塊體尺寸愈大，沖蝕模擬所得之沖蝕啟動剪應力隨岩體模型之節理間距呈大幅度升高。沖蝕啟動後，應以超額剪應力為基準，比較不同節理間距對平均沖蝕下切速率之影響；剪應力作用於岩床面個別塊體之作用面積越大，平均沖蝕下切速率也因而越高。 3. 藉由改變塊體細長比，探討塊體形狀因素對沖蝕行為之影響時，當塊體為非等邊長之塊體，傾角方向與水流方向可能出現順流抑或逆流兩種情況。順逆流方向影響塊體脫離岩床之運動方式，因此沖蝕啟動難易程度與沖蝕啟動後之行為，皆受順逆流方向之影響。

During flood season, river discharge in Taiwan varies significantly, stream power sometimes can become very high. The outcrop in the western foothill of Taiwan contains mostly young formation with low resistance and poor cementation. Once the armor layer in river channel becomes eroded away, the exposed rock bed will be subjected to irrecoverable erosion. Intensive incision or widening of riverbed causes the instability of river channels and threatens the safety of cross-river or near bank structures. The erosive mechanisms of bedrock and alluvium are very different. Among various erosion mechanisms, plucking is often the dominant mechanism of bedrock erosion when rock bed is composed of broken rock masses. For this situation, the erosion rate can be very high and deserves attention. Geologic factors affecting plucking mechanism are primarily related to the characteristics of weak planes in the rock mass: including discontinuity, orientation, block size, and block shapes, among others. Unlike running laboratory erosion tests for natural rock specimen, the usage of numerical simulation for a physical problem has several advantages. For example, it is easy to control over affecting factors; one can run as many cases of simulations as he wish. Furthermore, numerical simulation is less time-consuming and the specimen is always reproducible. This thesis hence uses numerical simulation as a tool for “virtual erosion test” in place of physical rock-mass erosion tests in reach scale. The virtual specimen is assumed to be heavily jointed rock mass containing two sets of weak planes in two dimensions. The virtual specimen of rock mass is subjected to shear traction on the top to simulate bed shear stress from water flow. From a series of virtual erosion tests, the following are summarized. 1. Effect of joint orientation on the threshold shear stress for plucking is examined by varying the slope of the climbing plane heading the flow direction of rock block. It appears the threshold shear stress for plucking increases with the angle of climbing plane. Once the applied shear stress exceeds the threshold value, erosion behavior will be no longer controlled by the angle of climbing plane, the erosion rate is controlled by the gentlest slope in the rock block. 2. Large spacing between joints will form larger block size. The threshold shear stress obtained from virtual erosion tests increases significantly with increasing joint spacing. Erosion rate is compared on the basis of excess shear stress (the amount of shear stress exceeds the threshold value). It appears the area of exposed plane to bed-shear traction affects the erosion rate of jointed rock mass. 3. The effect of block shape on plucking behavior is examined by varying the ratio of spacing of two joint sets. Slender rock blocks can be dipped along or against the flow direction. The angle between the dip angle and the flow direction affects the kinematic of blocks departing from the rock bed; hence, it thus affects the threshold shear stress and erosion rate for rock-mass plucking.