應用灰色理論於磨潤試驗之研究

The Study of Applying Grey Theory to Tribology Experiment

陳昭男(Zhao-Nan Chen)

指導教授 ： 謝宜宸

國立虎尾科技大學/工程學院/機械與機電工程研究所/碩士(2013年)

摘要

當兩機械元件有相對運動時，必定受到外力且互相接觸這時必須考慮表面間之接觸性質，如磨潤、變形與接觸方式等現象。為了模擬接觸元件之運動情形，透過多功能磨潤試驗機來進行本實驗，目的在於分析點接觸元件經由改變不同的控制參數與電阻值之關聯性，在不同控制參數正向力、側向力和轉速下觀察其變化的差異。
　　本文主要運用灰色系統理論進行磨潤分析研究，並經由實驗所得之數據，以灰關聯分析正向力、側向力和轉速下與電阻的關聯性分析，再建構灰預測GM（1,1）模型和RGM（1,1）模型，對不同的控制參數下電阻之預測。並以殘差檢驗以此檢測預測模型的精確度，是否達到準確的效果，進而了解接觸元件作動時的磨潤特性影響。
　　本研究所建構出灰關聯分析及灰預測理論GM（1,1）預測模型及RGM（1,1）模型，其結果發現：
一、經由灰關聯分析結果顯示，在不同的控制參數下，正向負荷的參數變動影響電阻值最大，側向負荷的參數變動影響電阻值其次，轉速負荷的參數變動影響電阻值最小。
二、灰關聯分析結果發現當正向力參數增加時，電阻值上升，磨擦損失增加。
三、GM（1,1）模型精確度均於94.03％以上，RGM（1,1）模型精確度均於94.82％以上，驗證此預測模型確實可以用於磨潤分析上。
四、RGM（1,1）模型預測的精準度比GM（1,1）模型預測出的精準度更高，RGM（1,1）模型更適合此分析之預測。
五、對預測而言，各種預測方法均有其適用的特點，所以預測方法的精確度不能全然做絕對的比較。若能找出不同預測模型並且符合決策的需求，所建立出來的預測模型，將具備更高之實用性與效益。

關鍵字

磨擦 ； 油膜 ； 電阻值 ； 灰色理論 ； 灰關聯 ； 灰預測

並列摘要

When there is relative motion between two mechanical elements, they will be in contact and affected by other physical forces. Hence, we should take the characteristics of their contact surfaces into consideration, such as tribology, deformation and the way they are in contact. To simulate the motion of the mechanical elements, multi-function tribology test machine was used in this research. By manipulating the relation between different control parameters and resistance values, the goal was to analyze their impacts and to observe the changes under different control parameters including normal forces, lateral forces and rotational speeds.
　　The study applied Grey Systems Theory to tribology test and used Grey Relational and the values of the experiments to analyze the relations among resistance and the normal force, lateral force and rotation speed. This research also built Grey Prediction GM (1,1) model and RGM(1,1) model for the prediction of resistance in different control parameters. Furthermore, Residual Checking was applied in order to examine the accuracy of the prediction models and their effects, in this way, aiming to understand the influence of tribology in the contact motion of the elements.
　　According to Grey Relational analysis and Grey Prediction theory GM (1,1) prediction models and RGM (1,1) models, the following were the findings.
(1)The result of Grey Relational analysis indicated that the parameters of normal load had the greatest effect on the resistance values, the influence of lateral load parameters on resistances decreased, and the rotational speed load had the smallest impact under different control parameters.
(2)The result of Grey Relational analysis indicated that the resistance values and the loss of friction increased when raising the parameters of normal force.
(3)The precision of GM (1,1) model was above 94.03%.The precision of RGM (1,1) model was above 94.82%. Consequently, this prediction model was appropriate for tribology analysis.
(4)The precision of RGM (1,1) model prediction was better than GM (1,1). Therefore, RGM (1,1) was more suitable for this prediction.
(5)In terms of prediction, every method has its advantage, and its precision should not be taken to a compatible comparison. If different prediction models can be found to meet the needs of various strategies, what is built up will have higher efficiency and practicability.

並列關鍵字

friction ； oil film ； resistance value ； Grey Theory ； Grey Relational ； Grey Prediction

參考文獻

[21]游凱文，2012，“ 側向力對接觸面積增長與接觸角影響研究 ”，虎尾科技大學機械與機電工程研究所學位論文。

[20]孫毅興，2010，“ 利用碎形理論建立量測訊號即時監測技術與磨潤行為關聯性之研究 ”，成功大學機械工程學系碩博士班學位論文。

[11]黃彥豪，2011，“ 以表面接觸電阻與史崔拜克曲線監測曲手軸承試驗機潤滑狀態之可行性研究 ”，崑山科技大學機械工程研究所論文。

[12]廖培凱、施鴻培、林明長(2009)，乾磨耗滑動條件下PEEK及UHMW-PE之磨潤特性研究，先進工程學刊4卷2期P.127 -132。

[4]林欣瑜、陳右儒、谷祖賢、林鶴南、方友清、童山(2005)，以摩擦力顯微術研究硬膜材料奈米磨潤特性，科儀新知第二十六卷第五期 。