Final geometrical variation and deviation have often a negative effect on product functionality and aesthetics. In the automotive industry, design concepts are being evaluated and optimized to withstand variation in the early phase of product development. For this end, simulation tools are employed. Input to these simulations is requirements on parts and fixtures or measurements from previously manufactured parts. In the case of plastic components, parts are often manufactured in the injection molding process. Here, different materials and process settings can make it difficult to predict deviation and variation based on similar parts. In order to perform accurate assembly variation simulation, part variation simulation need, therefore, to be included.
In this work a methodology is presented to simulate part and assembly variation, due to process noise, for plastic components manufactured in the injection molding process. The methodology is based on designed computer experiment and utilizes the concept of geometrical covariance and principal component analysis to relate process noise to variation patterns using regression analysis. Part and assembly variation are simulated combined using the distribution of these variation patterns.
The model used for part variation simulation has been verified against commercial injection molding software showing good agreement. An industrial case from the automotive industry is used to elicit the proposed methodology.
Manufacture 4.0 with LabVIEW and Arduino Mega for distributed cooling control of the DeMag 250 Ton machine injection mold
