3D Object Quality Prediction for Metal Jet Printer with Multimodal Thermal Encoder Network

With the advancements in 3D printing technologies, it is extremely important that the quality of 3D printed objects, measured in metrics including mechanical properties, and dimensional accuracies must meet the customer’s specifications. Various factors during Metal printing affect the printed part’s quality, including the powder quality, the printing stage parameters, the print part’s location inside the print bed, the curing stage parameters, and the metal sintering process. With the large data gathered from HP’s MetJet printing processes, AI techniques can be used to analyze, learn, and effectively infer the printed part quality metrics, as well as assist in improving the print yield. In-situ thermal sensing data captured by printer-installed thermal sensors such as the Heimann camera contains the part thermal signature of fusing layers. Such part thermal signature contains a convoluted impact from various factors that occur during metal printing. In this study, we used a multimodal thermal encoder network to fuse data of different natures including the video data and vectorized printer control data, and extract part thermal signatures with a trained encoder-decoder module. The pretrained model with efficient thermal feature extraction was then fused with printer control parameters for downstream tasks including part dimensional accuracy prediction and part porosity prediction. We explored the data fusing techniques and stages for data fusing, the optimized end-to-end model architecture indicated an improved part quality prediction accuracy.