Assessing Process Control in the Foundry

Abstract

Relying on scrap rates alone to assess process control in the foundry can be deceiving. Depending on the inspection method used, it is possible to sort grossly non-conforming parts, while still allowing a wide window of acceptance, masking a process that is out of control. Modern methods of acoustic resonance testing (ART) allow for detailed data logging of both part weight and resonant frequency peak data. Monitoring these two factors at the lot level can give foundry managers insight into how consistent their process is both within a single date code and across multiple date codes. The information collected during quality testing can and should be combined with statistical information gathered at the front of the foundry (design, tooling, molding, inoculation, pouring, cooling, and shakeout) and used to adjust the process to achieve smaller data spreads and more consistent and repeatable process control. Once a process is deemed to be under control, consideration should be given to monitoring the defined specification limits and subsequent control limits using statistical trend analysis as a means for ensuring a process is not approaching a condition in which the opportunity for quality issues can arise.