Statistical Process Monitoring With MTConnect

Abstract

Statistical Process Control (SPC) techniques are used widely in the manufacturing industry. However, it is sometimes observed that a deviation that is within the acceptable range of inherent process variation does not necessarily conform to specifications. This is especially true in the case of low volume; high precision manufacturing that is customary in aerospace and defense industries. In order to study the limitations posed by conventional SPC techniques in such manufacturing environments, a study was undertaken at TechSolve Inc., Cincinnati to develop a standalone SPC tool. The SPC tool so developed effectively communicates with an on-machine probe and analyzes the collected data to carry out a statistical analysis. MTConnect, a new-generation machine tool communications protocol, was used in developing the communication interfaces with the on-machine probe on a Computer Numerical Control (CNC) machine. The XML (eXtensible Markup Language) code used to extend the MTConnect schema to include the data obtained from the probing routines is also presented. The statistical analysis was developed as a Graphical User Interface (GUI) in LabVIEW. The statistical analysis was carried out as a case study by producing a widget. Real machining was carried out to produce 48 of these widgets using a combination of end mills and face mills. The data obtained during the subsequent quality testing was used to carry out the statistical analysis. The limitations of conventional SPC techniques during the developmental and analytical phases of the study are discussed. The presence of a chip during an on machine probing routine, the variations due to disparities in tool macro geometry, and the demand for conformance to requirements are studied in the view of a statistical process monitoring standpoint. Various alternatives are also discussed that aim to correct and improve the quality of machined parts in these scenarios.