本論文評估節理粗糙度係數（Joint Roughness Coefficient， JRC）是否足以作為單一描述參數，用以預測光滑至微粗糙岩石節理中應力依賴之力學孔隙（mechanical aperture， E）與水力孔隙（hydraulic aperture， e）。為釐清此問題，本研究建立一套可控制之實驗架構，使用具有可重複表面組態、吻合與不吻合條件，以及目標粗糙度範圍 JRC = 3.5 至 5.8 之 3D 列印圓柱狀節理類比試體進行測試。力學孔隙 E 係以節理體積法量測，水力孔隙 e 則由 YOKO2 系統之穩態氣體流試驗求得。本研究考量兩種端元流動組態，包含流體橫越粗糙波紋之橫向流，以及流體沿粗糙波紋方向流動之平行流。針對平行流系列，另使用感壓薄膜量化應力依賴之表觀接觸面積演化。 研究結果顯示，3D 列印節理類比試體架構能夠重現預定之節理幾何形貌，並產生具重複性且具物理意義之應力依賴孔隙變化趨勢。在本研究測試之粗糙度範圍與實驗室條件下，節理吻合程度對孔隙演化具有一階控制作用。吻合節理通常呈現較小孔隙、較規律之閉合行為，以及較明顯之表觀接觸發展；相較之下，不吻合節理則保留較大且更不均質之孔隙空間。單獨以 JRC 並無法充分整理或解釋所觀察到之孔隙反應。對於吻合節理，即使試體具有相近之 JRC 值，當其表面組態不同時，仍可能呈現不同之水力反應。對於不吻合節理，JRC 之影響較為明顯，但其作用仍受流動方向與殘餘流通通道連通性所調節。 ΔE-e 關係及其負斜率參數 -C 提供了一項有用之診斷指標，可用以描述水力孔隙對力學閉合之敏感性。直接 E-e 關係與 E/e-e 關係進一步指出，在本研究測試條件下，僅以 JRC 為基礎之關係式無法將力學孔隙唯一轉換為水力孔隙。進一步以 JRC、代表性表觀接觸條件描述參數，以及無因次化力學勁度項對 -C 進行之探索性結構參數化，可支持上述詮釋；然而，此參數化應被視為診斷性描述，而非已驗證之預測方程式。 整體而言，JRC 仍可作為描述粗糙度大小之一階指標；然而，對於本研究所測試之光滑至微粗糙節理，其應力依賴孔隙行為仍需結合節理吻合程度、表面組態、表觀接觸演化、流通路徑連通性與流動方向共同詮釋。;This thesis evaluates whether the joint roughness coefficient (JRC) is sufficient as a standalone descriptor for predicting stress-dependent mechanical aperture， E， and hydraulic aperture， e， in smooth-to-slightly rough rock joints. A controlled experimental framework was developed using 3D-printed cylindrical joint analogues with repeatable surface configurations， matched and mismatched conditions， and a target roughness range of JRC = 3.5 - 5.8. Mechanical aperture was measured using the joint-volume method， while hydraulic aperture was obtained from steady-state gas-flow measurements using the YOKO2 system. Two end-member flow configurations were examined: transverse flow across the roughness corrugations and parallel flow along the corrugations. For the parallel-flow series， pressure-sensitive films were used to quantify stress-dependent apparent contact-area evolution. The results show that the 3D-printed analogue framework reproduced the intended joint geometries and generated repeatable， physically meaningful stress-dependent aperture trends. Within the tested roughness range and laboratory conditions， matedness exerted a first-order control on aperture evolution. Matched joints generally showed smaller apertures， more systematic closure behavior， and more extensive apparent contact development， whereas mismatched joints retained larger and more heterogeneous void spaces. JRC alone did not fully organize the observed aperture responses. In matched joints， specimens with similar JRC values could exhibit different hydraulic responses when their surface configurations differed. In mismatched joints， the influence of JRC became more apparent， but remained conditional on flow orientation and residual channel connectivity. The ΔE-e relation and its negative-slope parameter， -C， provided a useful diagnostic measure of hydraulic sensitivity to mechanical closure. The direct E-e and E/e-e relations further indicated that mechanical aperture could not be uniquely converted into hydraulic aperture using a JRC-based relation alone under the tested conditions. The exploratory structural parameterization of -C using JRC， a representative apparent contact-condition descriptor， and a dimensionless mechanical-stiffness term further supports this interpretation， but should be regarded as diagnostic rather than predictive. Overall， JRC remains useful as a first-order roughness descriptor; however， stress-dependent aperture behavior in the tested smooth-to-slightly rough joints requires interpretation together with matedness， surface configuration， apparent contact evolution， flow-path connectivity， and flow orientation.
