An Investigation on Impact of Machine Learning in Additive Manufacturing

Abstract

One of the final phases in the product design process is prototyping or model creation. It is beneficial in the conception of a design. A model is produced and evaluated on a regular basis prior to the start of complete assembly. Historically, manual prototyping was employed to create a prototype. Additive manufacturing is a buzzword in the industrial and manufacturing industries. Initially, the CAD model of the components for the product was created in modeling software according to the specifications. Following the creation of a CAD model, the model is sliced by parallel planes equal to the layer thickness. As a result, the edges of these slices are quite sharp and squared, like a stair effect. These three-dimensional models will now be broken into small two-dimensional objects called slices. Simply said, a complicated three-dimensional problem has been reduced to a set of two-dimensional difficulties. These small two-dimensional files are known as STL files, and they are sent by tessellating the geometric three-dimensional model. Different surfaces of a CAD model are piecewise approximated by a sequence of triangles in tessellation, and the coordinates of triangle vertices and their surface normal are recorded. The predictive nature of various machine learning algorithms makes them the best instrument for dealing with additive manufacturing challenges. Machine learning techniques are capable of evaluating previous data and predicting future outcomes based on that analysis. This article discusses machine learning applications in additive engineering. © 2022 Elsevier B.V., All rights reserved.