設備非預期停機之研究-以薄膜電晶體液晶顯示器JI製程為例

The Study of Equipment Unexpected Downtime - A Case Study of JI Process in TFT-LCD Manufacturing

彭昱儒(YU-RU PENG)

指導教授 ： 鄭純媛 ； 孫德修

朝陽科技大學/理工學院/工業工程與管理系/碩士(2017年)

摘要

針對自動化生產設備，若發生非預期停機或異常停機時，可能導致生產線停擺、系統停止動作等問題，造成生產週期拉長、產品品質不良等，都使企業的營運成本大幅上升。本研究以薄膜電晶體液晶顯示器JI製程為例，從該製程設備的歷史停機資料，分別找出機構性與系統性發生異常停機嚴重的部位以及現象，並針對機構性與系統性異常停機部位之主要現象，依據其失效特性，探討失效次數曲線屬於單峰或多峰、檢定失效分配是否具有週期性以及其為單一或混合型失效機率分配，並建構一套建立較適配失效模式的流程與步驟。
本個案研究結果發現，機構性異常停機為混合型失效機率分配，其失效模式的第一段時間區間屬於機構磨耗或老化較緩慢的狀態，在第二段時間區間呈現機構磨耗或老化較快的狀態；系統性為單一失效機率分配，其失效模式找出(1)系統屬較成熟期，其失效率穩定地微緩漸增；(2)系統屬嬰兒期—失效率漸減型。
本研究所獲得成果，不僅提供個案公司作為後續找出真正停機的原因，更是提供個案公司在未來規劃保養模式或決策的參考，以及更新軟體時，對於經常失效的現象，找出有效的解決對策。甚至可作為未來發展自動修復停機機制的基礎。

關鍵字

非預期停機 ； 失效機率模式 ； 曲線擬合

並列摘要

For the production line with automatic equipment, it may cause production line breakdown, system stop action and other problems if an unexpected (or abnormal) downtime is occurred and then it results in longer production cycle time, less output quantity, and more operating costs. In this research, the unexpected downtime problem of the equipment of JI Process in a TFT-LCD Manufacturing factory is used as a case to study the categories of unexpected downtime and the most frequently occurred categories of the unexpected machine down. From the machine historical data of this case study, the downtime can be divided into normal and abnormal downtime according to the features of the downtime. For the severe abnormal downtime, this research aims to find out the critical failure parts and their major failure phenomena in both the mechanical and systematic failures. Then, for the critical parts and major failure phenomena, the failure characteristics and properties are studied and the failure modeling procedure by finding a more suitable fitted failure distribution to the real failure data is established.
It can be found from the research results that the failure models due to mechanical faults usually have mixed failure distribution models with lower aging status in each beginning period and with faster age status in each later period; the failure models due to system faults normally has single failure distribution with either a slightly increasing failure rate (systems in mature phase) or a decreasing failure rate (systems in infantile phase). The results of this research can provide the valuable information and models for the case company not only for figuring out the true causes of failure, but also for the decision-making in planning maintenance models. Furthermore, the results of this research can be the basis of developing the intelligent preventive maintenance models for the automatic manufacturing equipment.

並列關鍵字

unexpected downtime ； failure data ； fitted failure model

參考文獻

1.Ebeling, C. E., An Introduction To Reliability and Maintainability Engineering, McGraw-Hill companies, London, pp.23-40(1997).

2.Pulcini, G., “On the overhaul effect for repairable mechanical units: a Bayes approach”, Reliability Engineering and System Safety, 70: pp.85-94 (2000).

3.Rausand, M. and Hoyland, A., “System Reliability Theory: Models, Statistical Methods, And Applications”, Wiley-Interscience, pp.644 (2004).

4.Schittkowski, K., “EASY-FIT: a software system for data fitting in dynamical systems”, Springer-Verlag, pp.153-169 (2002).

5.Wiley, E. G., and Grosvenor, R. I., “Failure Data Analysis: the Application of Excel Spreadsheet Add-Ins as an Aid to Curve Fitting”, Quality and Reliability Engineering International Qual. Reliab. Engng pp.81-89 (2000).