| 摘要: | 岩體中的節理面構造不止影響岩體的本身材料強度，節理面粗糙度係數(JRC)也會影響岩體整體的剪力強度。Barton(1977)提出以視覺對照法判斷JRC，惟此判斷法較為主觀，近年來已有許多學者提出利用數值量化的方式評估JRC。為了更瞭解節理面粗糙程度與 JRC 值的關聯，魏培杰(2015)、廖明傑(2017)、羅宇軒(2018)蒐集並數位化 136 條岩石節理剖面，針對剖面高程差之標準差對其進行統計分析，迴歸出標準差與 JRC 的關係式。
本研究利用其關係式隨機產生 JRC=3、JRC=10、JRC=19.6 之剖面，利用 3D 列印技術列印出模型，並採用石膏與石英砂的複合材料灌製翻模符合指定節理面粗糙度的人造軟岩。上述人造節理面試體皆進行下列兩個步驟：(1)利用文獻中的四種 JRC 估算法驗證是否接近指定 JRC。(2)在正向力為0.5MPa、1.0MPa、 2.0MPa 下進行直剪試驗，以了解人造軟岩之節理試體在給定 JRC 下，其剪力強度是否亦符合 Barton 建議之公式。
研究結果顯示：(1)翻模後的人造軟岩節理面，其剖面 JRC 相較於指定JRC 更小，其原因可能與石膏試體於養治過程之收縮有關；(2) 以Barton經驗公式，並依試體真實 JRC 以及其單壓強度，推估之剪力強度與實驗之尖峰剪力強度差異百分比為 20%以內，顯示其人造軟岩之節理面剪力強度可大致以公式推估。(3)試體受剪後之破壞面積比與正向應力為正相關，試體JRC 對其較無影響。(4)低粗糙度試體在受剪後粗糙度會提高，高粗糙度試體受剪後其粗糙度會降低。
本研究也利用 PFC 軟體進行人造軟岩之單壓模擬及直剪模擬，以更進一步分析直剪試體之微觀力學行為，以單壓試驗為數值模型驗證之目標，為數值模型取得校正參數，並以此參數進行人造軟岩節理面之直剪模擬。
數值模擬結果顯示：(1)單壓模擬得到之結果與物理實驗高度吻合，包含其勁度、波松比以及尖峰強度。(2)以單壓模擬所得之數值參數進行直剪試驗模擬，直剪模擬結果之尖峰剪應力與勁度皆遠大於物理實驗，且破壞產狀與物理實驗不符。(3)以改善後之數值參數，其直剪模擬之尖峰剪應力及勁度已大致符合物理實驗，此結果亦顯示不同破壞機制之模擬，其參數亦可能有所不同。;The mechanical strength of the rock mass is affected by the joint surface structure and the joint surface roughness coefficient (JRC). Barton (1977) proposed an empirical chart to estimate the JRC by visual classification, but this method seems subjective. In recent years, many researchers have proposed using analytical methods to evaluate JRC more objectively. To better understand the relationship between the roughness of the joint surface and the JRC value, prior researchers in our group collected and digitized 136 rock joint profiles and performed statistical analysis on the standard deviation of the profile elevation difference. As a result, a relationship between the standard deviation and JRC profile height difference was proposed and proved valid.
In this study, the joint profiles of JRC=3, JRC=10, and JRC=19.6 were randomly generated by the proposed equation, and the cast model was printed using 3D printing technology.In addition, a large amount of artificial gypsum soft rock specimens was prepared by pouring the material into the model. We used these artificial soft rock jointed specimens to perform the following tasks: (1) Use the existing JRC estimation methods to verify the jointed specimen with the specified JRC values. (2) Carry out the direct shear tests under the normal stresses of 0.5MPa, 1.0MPa, and 2.0MPa.
The research results showed that: (1) the actual JRC of the artificial soft rock joint surface after curing is much smaller than that of the specified JRC (2) the peak shear strength estimated by the Barton empirical formula and the direct shear test is comparable. The percentage difference is within 20%. (3) The failure area ratio of the specimen after shearing is positively correlated with the normal stress, and the JRC of the specimen does not affect the failure area ratio. (4) The roughness of the low-JRC specimen is higher after shearing, and the roughness of the high-JRC specimen is reduced after shearing.
This study also uses PFC software to simulate artificial soft rock for uniaxial compression and direct shear tests to analyze the micro mechanical behavior. In addition, the physical uniaxial compression test was used to verify the numerical parameters in the PFC model, and the direct shear simulation was carried out with these parameters.
The numerical simulation results show that: (1) The results obtained from the uniaxial compression simulation are highly consistent with the physical experiments. (2) The direct shear simulation was first carried out using the verified parameters by a uniaxial compression test. As a result, the peak shear stress and stiffness of the simulation results are much larger than those of the physical experiment, and the failure conditions of the specimen are also inconsistent with the physical experiment. (3) The numerical parameters were modified to improve the simulation, as mentioned in earlier results. The peak shear stress and stiffness of the direct shear simulation are roughly in line with the physical experiments. |
