Identifying Unnatural Variation in Precision Rotational Part Manufacturing

Abstract

In the manufacturing industry, it is well known that in-process variation is a major contributor to poor quality products. In order to fabricate a precise part, the source of unnatural variation (UV) needed to be properly identified, monitored and controlled while the process is running. In relation to this issue, this study aims to identify the error root causes of UV in bivariate process associated with statistical process control (SPC) chart patterns. In research methodology, in-process variation in manufacturing roller head component was discussed systematically based on real product of roller head, computer aided design (CAD) and statistical process control (SPC) chart patterns. Initially, the CAD software was used to model a precise rotational part, and to analyse the cause of UV. Then, the programming software was used to generate the artificial SPC data streams based on an established mathematical model. Data generation also involved linear correlation between two dependent variables (bivariate). The outcome of this study would be helpful for industrial practitioners as a database when applying SPC for monitoring bivariate process.