ORALS

Uncertainty quantification (UQ) in metallic additive manufacturing (AM) has attracted tremendous interests in order to dramatically improve product reliability. Model-based UQ, which relies on the validity of a computational model, has been widely explored as a potential substitute for the time-consuming and expensive UQ solely based on experiments. However, its adoption in practical AM process requires the overcoming of two main challenges: (1) the inaccurate knowledge of uncertainty sources and (2) the intrinsic uncertainty associated with the computational model. Here we propose a novel data-driven framework to tackle these two challenges by combining high throughput physical simulations and limited experimental data. We first construct a machine learning (ML) model trained by high throughput physical simulations, for predicting the three-dimensional (3D) melt pool geometry and its uncertainty with respect to AM parameters and uncertainty sources. We then employ a novel sequential Bayesian calibration method to perform parameter calibration and model correction, by using experimental data from AM-Bench of National Institute of Standards and Technology (NIST). The application of the calibrated melt pool model to UQ of the porosity level, an important quality factor, of AM parts, demonstrates its potential use in AM quality control. The proposed UQ framework can be generally applicable to different AM processes, towards physics-based quality control of AM products.