Application Of Machine Learning Models To Predict Warping Of Plastic Automotive Parts

In order to manufacture plastic parts with complex geometries on a large scale, injection molding is one of the most widely used manufacturing processes. One of the criteria commonly considered when developing new products is dimensional quality, a factor directly associated with reducing part warpage. In the automotive industry, the manufacture of parts requires high dimensional precision, as many components have an aesthetic purpose or require perfect fits. Therefore, this work employs experimental designs, numerical simulations, statistical methods, and machine learning models to predict the warpage of an automotive cup holder. The polymers used were polypropylene (PP) and acrylonitrile butadiene styrene (ABS). The results indicated that the regression models developed for predicting warpage performed better with the ABS data. Regarding the classification models, both achieved an accuracy rate exceeding 90%. These findings provide useful tools during the mold tryout phase, helping to reduce time, cost, and material waste.