利用逆向模擬可靠度成長設計與DMAIC/DMADV相依模式之研究

The Research on Using the Inverse Simulation Reliability Growth Design and the Dependent Model of DMAIC/DMADV

關季明(Chi-Ming Kuan)

指導教授 ： 鄭春生

元智大學/工程學院/工業工程與管理學系/博士(2012年)

摘要

可靠度工程是確保產品研發能夠符合顧客需求的保證之程序與方法。六標準差為產品設計與品質改善之程序與方法。兩者為產品研發設計保證所利用之利器。然而，以往利用可靠度工程與六標準差改善及設計之售後品質仍具許多設計相關失效，造成顧客使用不滿意與市場流失。本研究衍生研究確保前述不足的改善以降低相關設計失效，研究提出的新方法為逆向模擬可靠度設計、早期監控與預測可靠度成長方法, 以及建立六標準差專案應用DMAIC/DMADV相依模式。

本研究對以往產品開發進行可靠度設計程序探討其效果，雖能完成產品開發。然而，通常在研發階段產生較多設計變更次數，且需依靠正規決算試驗方能驗證是否達成目標，如此可會使研發時間長，成本增加。本研究產生逆向模擬方法，探討電子產品之功能變異影響最大的溫度規格，以實際應力先以確定性分析訂定初步設計溫度規格，再利用逆向模擬方式將設定之初步溫度規格進行實際使用氣候溫度模擬確認生存機率 (survival rate) 高之溫度，方能確認為細步設計規格。當進行細步設計時，依可靠度配置而選用之零件溫度規格，以逆向失效樹分析，將實際使用氣候溫度作為虛擬感測器 (virtual sensor)，對選用之零件溫度規格模擬生存機率，如生存機率低於1時，則顯示零件選用不適當，可能導致系統不安全，可立即修正零件溫度規格直到生存機率近於1，方能保證此系統使用時作用安全，不易失效。

本研究發展早期監控可靠度成長模式，是考慮許多研究利用TAAF (test, analysis and fix) 計畫與DUANE模型作為進行可靠度成長的方法，在產品研發階段累積及分析相關可靠度試驗數據，即時監控產品的平均失效間隔時間 (mean time between failure, MTBF)，並追蹤產品可靠度成長狀況，使得產品在研發完成時可達到可靠度目標驗證。然而，過去進行這種可靠度成長試驗之前，必須先完成環境應力篩選 (environmental stress screening, ESS) 試驗，但未使用其數據與分析改善，而要重覆實施其他驗證試驗，此作法將耗費時間與成本。本研究提出一項可靠度驗證程序，首先在ESS試驗時，即進行TAAF以除去失效真因，再以修正之DUANE模型估計產品之MTBF，此方法以蒙地卡羅模擬來驗證多數研發專案之實施結果，證實可以確保達成可靠度目標之有效性。

本研究發展六標準差專案之DMADV為主軸與DMAIC相依併用之模式，是探討以往這兩項改善程序均為獨立專案實施時之效果，雖有成效互補性。然而，在DMAIC程序階段中無法全程滿足改善，則可能需要涉及流程重新設計方能完成改善，如拘泥於各自獨立程序可能產生之效益並非最佳，將耗時又耗成本，故本研究將DMAIC與DMADV過程成為相依模式交叉運用，建立一個DMADV為主軸，DMAIC為輔之相依模式，證實可得到極佳改善效果，並能縮短專案時間，以個案研究，探討一項在進行DMAIC之分析階段後，於改進階段無法達成改善目標，以相依方法執行後，確實改善產生六標準差專案最佳效益。

關鍵字

可靠度設計 ； 失效樹分析 ； 可靠度成長 ； 六標準差設計

並列摘要

The reliability engineering is the process and method of product research and development for fulfilling customer needs. The six sigma activities are the process and method of product design and quality improvement. These two methods are useful tool for conforming customer requirements during product design. However, a lot of design related failures still happened after applying the reliability engineering and six sigma improvements. This research proposed the methods to reduce failures through inverse simulation of reliability design specification, the early monitoring and forecasting the reliability growth method, and DMAIC/DMADV dependent model.

The method of the inverse simulation of reliability design specification is considered due to the reliability design often have many design changes during product development phase. However, reliability design requires deterministic analyze and long testing time test to demonstrate design goal (lower failure rate) which can be time consuming and costly in product design. This Research probed into electronic product by analyzing with deterministic temperature specified specification that customer required. And, the preliminary design temperature specification is stipulated. Then the inverse simulation method is applied to the imitating stipulated temperature specification in order to use actual climate temperature to verify the high survival rate at certain temperature. The designer will confirm the detail design specification to implement design. For the detailed design stage, dispose the goal in accordance with the reliability and select the part which accords with the determined temperature specification as before. During this period, this research will set up the inverse fault tree analysis and will apply the virtual sensor device to check the selected part temperature by simulation. If the survival rate is lower than 100%, it reveals the part is selected incorrectly and the system is unsafe. Then designer will revise the part temperature specification, until survival rate approaches to 100%. It will guarantee the systematic function of the products is safe.

This research considers early monitoring and forecasting the reliability growth method due to most related studies have focused on the reliability growth methods of the TAAF ( test analyze and fix ) program and Duane model. In the product design phase, appling test data to estimate and monitor the mean time between failure (MTBF) can track product reliability growth and complete reliability validation upon the completion of a design. In the past, the growth validation testing was executed after the environment stress screening. However, it is time consuming and costly. In addition, it can only detect workmanship not design problems. But, in practice, environment stress screening can identify many design problems during the design and development phase. Such problems would not occur in subsequent tests if they are removed in early phase. This research has proposed a new validation procedure. TAAF will be applied to find the root causes of failures during environment stress screening. MTBF are estimated by using the modified Duane model.

The DMAIC/DMADV dependent model builds the transforming procedure of the six sigma. These two activities of quality improvement are implemented independently in the project. However, it is unable to attain the objective during DMAIC procedure. It will further transfer to DMADV to finish project. This research model uses both DMAIC and DMADV alternately with DMADV as the core and combined with DMAIC. This model can assure the effectiveness of six sigma project.

並列關鍵字

esign for reliability ； fault tree analysis ； reliability growth ； design for six sigma

參考文獻

10. Cheng, C. S. and Kuan, C. M., 2006, Reliability Specification Deployment in DFSS, Proceedings of the Asia network quality congress.

1. Agarwal, H., Mozumder, C. K., Renaud, J. E. and Watson, L. T., 2007, An inverse-measure-based unilevel architecture for reliability-based design optimization, Structural and Multidisciplinary Optimization, 33(3), pp. 217-227.

2. Al-Kuwaiti, M. H., Kyriakopoulos, N. and Hussein, S., 2008, Towards a standardized terminology for network performance, IEEE Transactions on Reliability, 57(2), pp. 267-271.

3. Al-Mishari, S. T. and Suliman, S., 2008, Integrating Six-Sigma with other reliability improvement methods in equipment reliability and maintenance applications, Journal of Quality in Maintenance Engineering, 14(1), pp. 59-70.

4. Antony, J., Six Sigma: A strategy for supporting innovation in pursuit of business excellence, International Journal of Technology Management, 37(1-2), pp. 8-12.