Design of Manufacturing Process of Oxygen-Free High Conductivity Copper Using Mahalanobis-Distance Outlier Detection Method

Article Preview

Abstract:

The proper control of total impurities and oxygen contents of oxygen-free high conductivity (OFHC) copper prepared by vacuum high-frequency melting technique was studied using Mahalanobis-Distance (MD) outlier detection method as functions of raw material purities, vacuum pressure, melting temperature and holding time. The properties of vacuum-melted OFHC copper was examined by thermo-gravimetric analysis, differential scanning calorimetry, hardness test, macro and optical microstructure analyses and ultimate tensile test. In multivariate systems, the existence of outlier makes it difficult to analyze the system and oultier detection belongs to the most important tasks in experimental data analysis. Mahalanobis Distance is most commonly used as a diagnosis of existance of outlier in multivariate system. The relationship between experiment conditions and total impurities and oxygen contents can be defined with the regression analysis results. At this research, our desirable manufacturing conditions is to obtain the total impurities under 40 ppm and oxygen contents under 5 ppm. After this statistical approach, the suggested minimum maufacturing conditions are the purity of raw material was 4N, vacuum pressure was 10-1 torr, melting temperature was 1150°C and melt holding time was 20 minutes.

You might also be interested in these eBooks

Info:

Periodical:

Materials Science Forum (Volumes 544-545)

Pages:

965-968

Citation:

Online since:

May 2007

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2007 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

[1] Olin Brass: Understanding Copper Alloys, edited by J. Howard Mendenhall, A WileyInterscience Publication, John Wiley & Sons, NY (1980).

Google Scholar

[2] S. Aoki, Y. Shikano and K. Yajima: J. JRICu, Vol. 42, No. 1 (2003), p.21.

Google Scholar

[3] Y.O. Yoon, H.H. Jo, H. Cho, S.K. Kim and Y.S. Kim: J. Kor Inst. Met & Mater., Vol. 43, No. 3 (2005), p.255.

Google Scholar

[4] P. Filzmoser: Information on http: /www. statistik. tuwien. ac. at.

Google Scholar

[5] D.C. Montgomery: Design and Analysis of Experiments, 3 rd Edition, John Wily & Sons, (1991).

Google Scholar

[6] Multiple regression: Information on http: /www. statsoft. com/textbook/stmulreg. html.

Google Scholar