本研究針對既有滾子凸輪研究較少整合CAD運動軌跡建模、槽寬比較、曲線修正與雙向動態驗證之問題，建立低背隙滾子凸輪之設計驗證流程。首先利用SolidWorks Motion取得滾子相對主動輪之運動軌跡，反推槽道中心螺旋曲線，再串接VBA、Mathematica與SolidWorks完成點位擷取、偏移計算及模型重建，並以ADAMS進行單向完整週期與雙向運轉模擬。原始曲線共取樣801點，平均相鄰點距0.3066 mm，最大局部弦高量級約1.3 μm，低於10 μm偏移修正尺度。結果顯示，單側名義間隙30 μm之W1006僅完成約10%理論輸出而判定失效；有效模型中，W1008相較W1010之傳動誤差角均方根值降低34.8%，輸出端等效背隙角降低40.1%，但W1010具有較低接觸負載，顯示槽寬設計存在傳動精度、背隙與接觸負載之取捨。固定10.10 mm槽寬並施加最大10 μm漸進式偏移後，P75之反向響應時間差由0.031 s降至0.012 s，降低61.3%，輸出端等效背隙角由4.446″降至0.522″，降低88.3%。結果顯示，在不進一步縮小槽寬之前提下，微量修正槽道中心曲線可有效改善低背隙反向響應；所建立流程可於實體加工前比較槽寬調整與曲線修正策略，並降低試作與修模需求，作為後續滾子凸輪槽道設計與動態驗證之參考。;This study develops an integrated design and verification process for low-backlash roller cams by combining CAD-based motion trajectory modeling， groove-width comparison， curve correction， and multibody dynamics simulation. SolidWorks Motion was used to obtain the roller trajectory relative to the driving cam and derive the groove center curve; VBA and Mathematica were then applied for point extraction， offset calculation， and model reconstruction， followed by ADAMS simulations. The curve was sampled at 801 points with a mean spacing of 0.3066 mm and a maximum local chord-height magnitude of approximately 1.3 μm， below the 10 μm correction scale. W1006， with a nominal single-side clearance of 30 μm， completed only about 10% of the theoretical output and was classified as invalid. Among the valid models， W1008 reduced the RMS transmission error by 34.8% and the equivalent backlash angle by 40.1% compared with W1010， whereas W1010 produced a lower contact load， indicating a trade-off among accuracy， backlash， and contact load. With the groove width fixed at 10.10 mm， the P75 correction reduced the reverse response delay from 0.031 s to 0.012 s and the equivalent backlash angle from 4.446 arcsec to 0.522 arcsec， corresponding to reductions of 61.3% and 88.3%， respectively. The proposed process enables groove-width and curve-correction strategies to be evaluated before manufacturing， reducing prototyping and rework requirements.
