Deep Learning for Distortion Prediction in Laser-Based Additive Manufacturing using Big Data

Abstract

Laser-Based Additive Manufacturing (LBAM) is a fabrication process that is a key aspect of Industry 4.0, which aims to employ many sensors for continuous process control. One current challenge in LBAM is the geometric inaccuracy of fabricated parts. To increase accuracy, accurate predictions of distortion are needed. Here we develop a novel Deep Learning approach that accurately predicts distortion well within LBAM tolerance limits by considering the local heat transfer for pointwise distortion prediction. Our Deep Learning approach not only gives highly accurate predictions but also fits into the Industry 4.0 framework of analyzing big data with many sensors.