臺灣博碩士論文加值系統

SKD11合金工具鋼是一種高碳高鉻合金工具鋼，這種鋼材經過熱處理後，硬度可達到HRC58-62，亦即它具有高強度、高硬度、高韌性、表面抗磨耗能力、較高的疲勞強度和抗變形等優越特性，在工業中，主要廣泛應用於製造沖壓模、塑膠模等多種模具。
超音波輔助銑削顯著改善表面粗糙度並減少切削力，但大多數研究僅集中於特定參數。此外，傳統的冷卻方法在提升加工效率和延長刀具壽命方面有其侷限性，特別是在高速切削條件下，冷卻效果難以應對加工過程中產生的大量熱量和摩擦。本研究透過綜合評估不同輔助和冷卻方式，制定更有效的切削策略。並利用超音波輔助的高頻振動與微小振幅產生的泵吸作用與中心出水冷卻精準噴射且吸入於主切削區域的特性相結合，達到更佳的冷卻與潤滑效果，為提升SKD11的切削性能開闢了新的思路。
本文研究不同輔助方式對SKD11銑削性能的影響，並找出較佳的輔助方式和切削參數。首先進行不同超音波輔助方式的比較，包括Z軸超音波、XY軸超音波和XYZ三軸超音波，並比較乾切削、外部冷卻和中心出水等不同的冷卻方式。實驗結果顯示，結合較佳的超音波輔助和冷卻方式能顯著提升切削性能並減少切削過程中的熱量產生。接著，運用田口法進行銑削實驗，選定較佳的超音波輔助、冷却方式以及切削製程參數。結果顯示，使用Z軸超音波輔助時，無論是在小切削深度低進給率還是大切削深度高進給率的情況下，均能顯著改善表面粗糙度，表面粗糙度比無輔助條件下降20%-31%。此外，中心出水冷卻方式能直接將冷卻液輸送至切削區域，有效降低切削溫度，並透過其高壓冷卻液打斷硬化後的切屑，促進其快速排出，進而大幅度改善表面粗糙度，與乾切削相比，其表面粗糙度降低15%-28%。
在本文應用多種機器學習模型，包括決策樹、隨機森林、支持向量機等8種基本分類模型，來預測刀具不同階段的磨耗。AdaBoost模型在眾多模型中表現最佳，顯示預測的準確性。最後，透過集成學習投票法的應用顯示，透過多模型的結合可以進一步提升預測精確度，為製造業提供更為穩定和高效的刀具管理方案。

SKD11 alloy tool steel is a high-carbon, high-chromium alloy tool steel. After heat treatment, this material can achieve a hardness of HRC58-62, meaning it possesses excellent properties such as high strength, high hardness, high toughness, surface wear resistance, high fatigue strength, and resistance to deformation. In the industry, it is widely used in the manufacturing of various molds, including stamping dies and plastic molds.
Ultrasonic-assisted milling significantly improves surface roughness and reduces cutting force, but most studies have focused only on specific parameters. In addition, traditional cooling methods have limitations in improving machining efficiency and extending tool life, especially under high-speed cutting conditions, where cooling efficiency is insufficient to handle the large amounts of heat and friction generated during the processes. This study comprehensively evaluates different assisted-machining and cooling methods to develop more effective cutting strategies. By combining the pumping effect generated by high-frequency vibrations and small amplitudes in ultrasonic assistance with the precise injection and suction characteristics of internal cooling in the main cutting area, better cooling and lubrication effects are achieved, opening new avenues for enhancing the cutting performance of SKD11.
This paper investigates the effects of different assisted methods on the milling performance of SKD11 and identifies the optimal assisted methods and cutting parameters. First, a comparison of different ultrasonic assisted methods was conducted, including Z-axis ultrasonic, XY-axis ultrasonic, and XYZ three-axis ultrasonic, along with a comparison of dry cutting, external cooling, and internal cooling. The experimental results show that combination of the optimal ultrasonic assistance and cooling methods can significantly improve cutting performance and reduce heat generation during the cutting processes. Subsequently, the Taguchi method was used to conduct milling experiments, by selecting the better ultrasonic assistance, cooling methods, and cutting process parameters. The results show that using Z-axis ultrasonic assistance, whether at low feed rates with small cutting depths or high feed rates with large cutting depths, can significantly improve the surface roughness, reducing surface roughness by 20%-31% compared to conditions without assistance. Additionally, the internal cooling method can directly deliver coolant to the primary cutting area, effectively reducing cutting temperature and constituting brittle chip. These chips may be broken off via the impact of high-pressure coolant, promoting their rapid removal from the workpiece, which in turn greatly improves surface roughness. Compared to dry cutting, surface roughness is reduced by 15%-28%.
In this paper, various machine learning models, including decision trees, random forests, support vector machines, and other basic classification models, are applied to predict tool wear at different stages during the milling processes. The AdaBoost model performed the best among the other models, showing the highest prediction accuracy. Finally, the application of ensemble learning voting demonstrated that the integration multiple models can further improve prediction accuracy, providing the manufacturing industry with a more stable and efficient tool management solution.

摘要.....i
Abstract.....ii
誌謝.....iv
目錄.....v
表目錄.....vii
圖目錄.....viii
第一章 前言.....1
1.1 研究背景.....1
1.2 研究動機及目的.....2
1.3 文獻回顧.....2
1.4 論文架構.....8
第二章 理論基礎.....9
2.1 刀具磨耗.....9
2.2 超音波輔助.....10
2.3 中心出水冷卻.....11
2.4 機器學習.....11
2.5 集成學習.....12
第三章 實驗規劃與方法.....14
3.1 工件與刀具.....15
3.2 實驗儀器設備與規格.....16
3.3 銑削實驗架構.....24
3.4 機器學習架構.....27
3.4.1 資料生成.....28
3.4.2 資料清理.....30
3.4.3 特徵提取.....32
3.4.4 資料切割與標準化.....33
3.4.5 特徵選擇.....33
第四章 結果與討論.....34
4.1 第一階段較佳超音波.....34
4.2 第二階段較佳超音波及較佳冷卻.....42
4.3 第三階段田口實驗直交表.....48
4.4 機器學習.....52
4.4.1 機器學習各模型之預測分數.....53
4.4.2 機器學習與集成學習比較.....57
第五章 結論.....58
參考文獻.....59

[1]Hui, Y., et al., 2019, "Milling tool wear state recognition by vibration signal using a stacked generalization ensemble model", Shock and vibration, vol 2019, pp. 1-16.
[2]Chuangwen, X., et al., 2018, "The relationships between cutting parameters, tool wear, cutting force and vibration", Advances in Mechanical Engineering, vol 10, no 1, pp. 1687814017750434.
[3]Karandikar, J., et al., 2015, "Tool wear monitoring using naive Bayes classifiers", The International Journal of Advanced Manufacturing Technology, vol 77, pp. 1613-1626.
[4]Mejbel, A., et al., 2021, "Improving the machined surface of AISI H11 tool steel in milling process", J. Mech. Eng. Res. Dev, vol 44, no 4, pp. 58-68.
[5]Tsai, M. Y., et al., 2016, "The machining of hard mold steel by ultrasonic assisted end milling", Applied Sciences, vol 6, no 11, pp. 373.
[6]Gao, H., et al., 2022, "Enhancement of machinability and surface quality of Ti-6Al-4V by longitudinal ultrasonic vibration-assisted milling under dry conditions", Measurement, vol 187, pp. 110324.
[7]Wang, C., et al., 2016, "Modeling and simulation of the high-speed milling of hardened steel SKD11 (62 HRC) based on SHPB technology", International Journal of Machine Tools and Manufacture, vol 108, pp. 13-26.
[8]Wang, C., et al., 2015, "Wear and breakage of TiAlN-and TiSiN-coated carbide tools during high-speed milling of hardened steel", Wear, vol 336, pp. 29-42.
[9]Gong, F., et al., 2019, "Wear and breakage of coated carbide tool in milling of H13 steel and SKD11 hardened steel", SN Applied Sciences, vol 1, pp. 1-12.
[10]Pu, Z., Singh, A., 2013, "High speed ball nose end milling of hardened AISI A2 tool steel with PCBN and coated carbide tools", Journal of Manufacturing Processes, vol 15, no 4, pp. 467-473.
[11]Gong, F., et al., 2017, "Fatigue failure of coated carbide tool and its influence on cutting performance in face milling SKD11 hardened steel", International Journal of Refractory Metals and Hard Materials, vol 64, pp. 27-34.
[12]Zhang, G., et al., 2022, "Effect of SiC nanofluid minimum quantity lubrication on the performance of the ceramic tool in cutting hardened steel", Journal of Manufacturing Processes, vol 84, pp. 539-554.
[13]Shinge, V., Pable, M., 2023, "Effect of nano-minimum quantity lubrication on cutting temperature and surface roughness of milling AISI D3 tool steel", Materials Today: Proceedings, vol 72, pp. 1758-1764.
[14]Taylor, C. M., et al., 2020, "Resource consumption and process performance in minimum quantity lubricated milling of tool steel", Procedia Manufacturing, vol 43, pp. 463-470.
[15]Ravi, S., Gurusamy, P., 2020, "Experimental investigation of cryogenic cooling on cutting force, surface roughness and tool wear in end milling of hardened AISI D3 steel using uncoated tool", Materials Today: Proceedings, vol 33, pp. 3314-3318.
[16]Jamil, M., et al., 2022, "A novel low-pressure hybrid dry ice blasting system for improving the tribological and machining characteristics of AISI-52100 tool steel", Journal of Manufacturing Processes, vol 80, pp. 152-160.
[17]Ghule, G. S., et al., 2022, "Experimental investigations on the ultrasonic vibration-assisted hard turning of AISI 52100 steel using coated carbide tool", Materials Today: Proceedings, vol 68, pp. 2093-2098.
[18]Chen, X., et al., 2015, "Cutting performance and wear characteristics of Ti (C, N)-based cermet tool in machining hardened steel", International Journal of Refractory Metals and Hard Materials, vol 52, pp. 143-150.
[19]Saito, H., et al., 2016, "Elliptical vibration cutting of hardened die steel with coated carbide tools", Precision Engineering, vol 45, pp. 44-54.
[20]Tapoglou, N., et al., 2017, "Investigation of the influence of CO2 cryogenic coolant application on tool wear", Procedia Cirp, vol 63, pp. 745-749.
[21]Pereira, O., et al., 2017, "Internal cryolubrication approach for Inconel 718 milling", Procedia Manufacturing, vol 13, pp. 89-93.
[22]Zhang, C., et al., 2016, "Effects of internal cooling channel structures on cutting forces and tool life in side milling of H13 steel under cryogenic minimum quantity lubrication condition", The International Journal of Advanced Manufacturing Technology, vol 83, pp. 975-984.
[23]Khosravi, J., et al., 2022, "High-speed milling of Ti6Al4V under a supercritical CO2+ MQL hybrid cooling system", Journal of Manufacturing Processes, vol 82, pp. 1-14.
[24]Twardowski, P., et al., 2023, "Monitoring and forecasting of tool wear based on measurements of vibration accelerations during cast iron milling", Journal of Manufacturing Processes, vol 95, pp. 342-350.
[25]Niu, B., et al., 2020, "Multisensory based tool wear monitoring for practical applications in milling of titanium alloy", Materials Today: Proceedings, vol 22, pp. 1209-1217.
[26]Mahmood, J., et al., 2022, "Accurate estimation of tool wear levels during milling, drilling and turning operations by designing novel hyperparameter tuned models based on LightGBM and stacking", Measurement, vol 190, pp. 110722.
[27]Zhou, H., et al., 2023, "Multi-condition wear prediction and assessment of milling cutters based on linear discriminant analysis and ensemble methods", Measurement, vol 216, pp. 112900.
[28]Xie, Y., et al., 2024, "Tool wear state recognition and prediction method based on laplacian eigenmap with ensemble learning model", Advanced Engineering Informatics, vol 60, pp. 102382.
[29]Huang, Z., et al., 2023, "Hybrid machine learning-enabled multi-information fusion for indirect measurement of tool flank wear in milling", Measurement, vol 206, pp. 112255.
[30]Çınar, Z. M., et al., 2020, "Machine learning in predictive maintenance towards sustainable smart manufacturing in industry 4.0", Sustainability, vol 12, no 19, pp. 8211.
[31]Sarker, I. H., 2021, "Machine Learning: Algorithms, Real-World Applications and Research Directions", SN Computer Science, vol 2, no 3, pp. 160.
[32]Sagi, O., Rokach, L., 2018, "Ensemble learning: A survey", Wiley interdisciplinary reviews: data mining and knowledge discovery, vol 8, no 4, pp. e1249.
[33]Xue, Z., et al., 2023, "Hierarchical temporal transformer network for tool wear state recognition", Advanced Engineering Informatics, vol 58, pp. 102218.
[34]Elbestawi, M., et al., 1991, "In-process monitoring of tool wear in milling using cutting force signature", International Journal of Machine Tools and Manufacture, vol 31, no 1, pp. 55-73.
[35]Kuljanic, E., Sortino, M., 2005, "TWEM, a method based on cutting forces—monitoring tool wear in face milling", International Journal of Machine Tools and Manufacture, vol 45, no 1, pp. 29-34.
[36]Gierlak, P., et al., 2017, "On-line manipulator tool condition monitoring based on vibration analysis", Mechanical Systems and Signal Processing, vol 89, pp. 14-26.
[37]Mao, X., et al., 2021, "Semi-random subspace with Bi-GRU: Fusing statistical and deep representation features for bearing fault diagnosis", Measurement, vol 173, pp. 108603.
[38]Qin, Y., et al., 2023, "Tool wear identification and prediction method based on stack sparse self-coding network", Journal of Manufacturing Systems, vol 68, pp. 72-84.