A Comparison of Statistical Process Control (SPC) and On-Line Multivariate Analyses (MVA) for Injection Molding

Abstract

Manufacturing process automation is often impeded by limitations related to automatic quality assurance. Many plastics manufacturers use univariate statistical process control (SPC) for quality control by charting the critical process states relative to defined control limits. Alternatively, principal component analysis (PCA) and projection to latent stuctures (PLS) are multivariate methods that measure the process variance by the distance to the model (DModX) and the Hotelling t-squared (T2) values. A methodology for robust model development is described to perturb the manufacturing process for process characterization based on a design of experiments; best subset analysis is used to provide an optimal set of regressors for univariate SPC. Four different statistical models were derived from the same data set for a highly instrumented injection molding process. The performance of these models was then assessed with respect to fault diagnosis and defect identification when the molding process was subjected to twelve common process faults. Across two hundred molding cycles, the univariate SPC models correctly diagnosed five of the twelve process faults with one false positive, detecting only eighteen of twenty four defective products while indicating two false positives. With the same molding cycles, PCA and PLS provided nearly identical performance by correctly diagnosing ten of the twelve process faults and detecting twenty three of the twenty four defective products; PCA indicated two false positives while PLS indicated only one false positive.

Mold Commissioning

“…However, the algorithm and data structures are complex and still subject to uncertainty in the models, y j = f(x i ), that relate the quality attributes to the processing conditions [26]. A practical approach to confirming the robustness of a newly commissioned mold is to perform a design of experiments (DOE) centered around the proposed operating conditions [27,28].…”

Meta‐analysis of thermal contact resistance in injection molding: A comprehensive literature review and multivariate modeling

“…Accordingly, using PCA with ALS would be expected to retain the TCR behavior reported in the literature while increasing the robustness by drawing on all observations. As typical in PCA, 38 we coded the data into a matrix, X, where the columns correspond to the different factors/responses, and each row represents an observation that may include missing data coded as either the empty set, [], or not a number, #NaN. Each column is centered by subtracting the mean of the observations in the column and normalized by dividing the standard deviation of the observations in the column.…”

An investigation of the utilisation of different data sources in manufacturing with application in injection moulding

“…The injection moulding process can be seen as a fast batch process (8−30 s), as each 'shot' yields time series data that is expected to follow a predetermined profile, see examples of injection pressure profiles in Figure 3. It has been well established that quality assurance in injection moulding requires the analysis of multiple process variables simultaneously as in the multivariate statistical process control (MPSC) (Ambrozic and Hutson 2006;Hazen and Kazmer 2007;Kazmer, Westerdale, and Hazen 2008).…”