Comprehensive Quality Investigations of Wire-feed Additive Manufacturing based on Machine Learning of Experimental Data
Year: 2021
Type: preprint
Abstract: Wire-feed Laser-based Additive Manufacturing has gained attention for years with its promises of high-level automation, reducing materials waste, overall costs, and large-scale volume production. However, comprehensive quality quantification and control are difficult to implement due to the limitation of dimensionality and cost of experimental data that traditional trial and error methods and computational methods can handle. To enable quality control of desired part geometry and microstructural properties of the part, this study proposes a comprehensive quality investigation framework based on learning from experimental data. Specifically, experimental data across various process variables and output properties in terms of bead overall quality, geometry (i.g. bead height, fusion zone depth, etc.), and microstructural properties are collected. The predictive process-property relations are then modeled with data-driven ML classification and regression models. The multiple properties of the printed bead are predicted on processing space based on a 3-dimensional contour map. The insights and impacts of process variables on bead morphology, geometric and microstructural properties are comprehensively investigated. The processing parameters leaned by the models can rapidly designing materials with desired functional properties.
Source: arXiv (Cornell University)
Authors: Sen Liu, Craig Brice, Xiaoli Zhang
Institution: Colorado School of Mines
Language: English
Subfield: Mechanical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3136674802