Process Parameter Optimization of Additively Manufactured Parts using Intelligent Manufacturing

Abstract

Additive manufacturing is an essential element in the manufacture of machinery. Over the years, researchers have planned to innovate with new ideas for improving laminated molding techniques. This article aims to optimize process parameters. Several machine-learning techniques were used to solve the problem, but they were tedious and time-consuming. The Azure ML database was used to mitigate these errors. This gave similar results without writing a massive line of code. The motivation of this thesis is to improve the tensile strength of objects by optimizing process parameters. Above all, the printer selection was made on a per-order basis. SLA and FDM printers are a hot topic in today's laminated modeling, so a detailed literature review was conducted. FDM printers are used for research work because SLA printers are costly, and the print quality is good. Tensile strength was evaluated in relation to infill density, infill pattern, layer height, and the number of perimeter walls. After measuring the force of each part, the data was uploaded to the Azure ML portal, a linear regression model was applied, a prediction engine was built, and multiple input parameters were defined to forecast different tensile strength values on the web to extend the service model.