Abstract
Researchers detail new data in Machine Learning. According to news reporting from Augsburg, Germany, by NewsRx journal ists, research stated, "Injection molding is a popular production process for sh ort fiber reinforced components. The mechanical properties of such components de pend on process-induced fiber orientations which are commonly predicted via nume rical simulations." Funders for this research include Hightech Agenda Bavaria, Bavarian State Govern ment. The news correspondents obtained a quote from the research from the University o f Augsburg, "However, high computational costs prevent process simulations from being used in iterative procedures, such as topology optimization or finding opt imal injection locations. We propose a fast approximation method that extracts n odal features and train a regression model to predict fill states, cooling times , volumetric shrinkage, and fiber orientations. The features are determined by s olving eikonal equations with a fast iterative method and computing spatial mome nts to characterize node-adjacent material distributions. Subsequently, we use t hese features to train feed forward neural networks and gradient boosted regress ion trees with simulation data of a large dataset of geometries. This approach i s significantly faster than conventional methods, providing 20x speed-up for sin gle simulations and more than 200x speed-up in gate location optimization."
NETL
NSTL