Process Modeling and Capability Feedback for Integrated Injection Molded Product Development

In order to limit new product development time and ensure quality, the engineer needs to understand the capability of the manufacturing process that will be used. Specifically, the engineer must be able to assess the feasibility of producing the parameters (i.e. dimensions and physical properties) of a design within their specified tolerances and understand how variation will affect the ideal function of the product. Concurrent engineering has taken a holistic approach to the problem by developing management methods to promote greater communication between designers and manufacturing engineers. However, this method of assessing process capability relies only on the "experience" of manufacturing engineers. A more quantitative method is needed. This thesis considers the needs of both design and manufacturing engineers with regards to understanding the capability of a manufacturing process. A methodology is presented that shows how an enterprise can improve its understanding of its manufacturing processes by effectively managing the information generated by these processes. It is proposed that feedback of production data to process modeling tools used in the design phase will satisfy this need. The application of this methodology will lead to continuous improvement of the capability of both the manufacturing process and the design process. These improvements will stem from a better understanding of how to apply manufacturing process models. Furthermore, engineers will be able to improve the management of tolerance allocation and process development efforts. The details of the implementation of this methodology are demonstrated through a case study involving injection molding. A model was created to enable design and manufacturing engineers to anticipate and track the variation of critical parameters, thereby promoting feedback. A commercial software package for the simulation of injection molding was evaluated through the use of designed experiments. These experiments were run on a production line and in simulation. The verification included investigations of shrinkage and warpage predictions. Finally, methods of feedback into first-order models were developed. Based upon the results of this study, a model usage strategy was formulated. Thesis Advisor: Kevin Otto 'Title: Assistant Professor of Mechanical Engineering