Geometric Variation Simulation for the Development of Products with Plastic Components

Licentiate thesis, 2012

Every manufactured product deviates from the intended product. Furthermore, in a series of products geometrical variation is always present. This variation will lead to functional and aesthetical variation of the product. Geometrical variation stems from variation in the manufacturing of parts and fixtures and variation in assembly operations. Also, functional variation is dependent on environmental variation during the user phase. The source of variation accumulates during production. Often, it is difficult or expensive to reduce the source of variation. Therefore, instead of diminishing the source of variation it is preferable to reduce the effect of this variation. This is the case in a robust design. In connection to geometrical variation a robust design is often related to the way parts are located to other parts in a product. It is in the early phase of product development that most design solutions are determined. Design changes that are introduced later in the product development process often generate a high cost. Hence, to gain confidence during the design phase, that critical measures and key characteristics are within their requirements for the manufactured product it is important to be able to predict product behavior. This is done using virtual product development tools and/or prototypes. Virtual tools have a lot of advantages in comparison to physical prototypes. It is, however, important that simulations are accurate and that all relevant phenomena are considered. In this work the focus is on variation simulation for plastics. The increasing environmental concern has led a number of industries to investigate the possibilities to reduce the weight of their products. The relative ease of the manufacturing techniques and their flexible physical properties have made plastics an attractive alternative. However, this leads to new challenges in virtual product development. In this thesis an interview study is performed that reports current issues and problems when simulating for robustness in plastic design. This led to a framework for robust plastic design where part-, assembly and functional assembly are considered as different levels of robustness. A procedure is proposed where commercial injection molding techniques are used in combination with statistical methods to predict part variation resulting from manufacturing variation. This is implemented to study the effect of manufacturing variation on assembly variation. Furthermore, a tool is developed to simulate thermal expansion in combination with assembly variation. A case study reveals that there are combined effects. Finally, a method to stabilize the numerical calculation of the thermal elasto-plastic problems is developed.