Multivariate regression modeling for monitoring quality of injection moulding components using cavity sensor technology: Application to the manufacturing of pharmaceutical device components
| Autor: | Magida Zeaiter, Wendy Anne Knight, Holland Simon Joseph |
|---|---|
| Rok vydání: | 2011 |
| Předmět: |
Engineering drawing
Engineering business.industry Design of experiments media_common.quotation_subject Final product Process (computing) Regression analysis Industrial and Manufacturing Engineering Quality by Design Computer Science Applications Control and Systems Engineering Modeling and Simulation Partial least squares regression Injection moulding Quality (business) Process engineering business media_common |
| Zdroj: | Journal of Process Control. 21:137-150 |
| ISSN: | 0959-1524 |
| Popis: | There is an increased demand within the moulding industry to improve the quality of moulded parts by maintaining consistent tolerances and overall dimensions. This interest is more important in areas of the moulding industry that are dedicated to pharmaceutical devices, where a quality by design approach is now expected to be adopted. A pharmaceutical device is an assembly of different plastic components which are manufactured by injection moulding; many have critical quality parameters which affect not only the device appearance but also more importantly its performance for drug delivery. Hence, the need of better understanding and control of injection moulding processes. This study presents the use of multivariate regression modeling approach to monitor the quality of the final product using cavity sensor technology (CST). The influence of the injection moulding process parameters on the quality of the final parts have been investigated using a design of experiment approach. The results demonstrate that the Partial Least Squares (PLS) regression model based on cavity pressure sensor data could be successfully used to monitor the quality (weight, dimensions) of the final product in plastic injection moulding. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect