Multi-response Optimization of 3D Printed Parts with Triangular Patterns Using Nonlinear Machine Learning Regressor Technique

Abstract

In the ever-changing field of additive manufacturing, this work implemented machine learning algorithm (nonlinear regression approach) to find out the optimal tensile strength and modulus of elasticity of 3D-printed ABS, PLA, and PLA + CF components using MATLAB R2021a. The work navigates the complex interactions among material type, orientation, and infill density using a complex experimental design matrix. With an impressive accuracy rate of more than 90%, the nonlinear model proves to be a useful tool for predicting mechanical properties in the context of 3D printing. The study uses the desirability technique as part of a multi-response optimization strategy to achieve optimal performance. By using this procedure, the best process parameters are identified, and the result is a composite desirability score of 0.84. An orientation fixed at 0�, a material type defined as PLA + CF, and an infill density set at 80% comprise the identified peak of efficiency. In addition, the ideal parameters for the responses of Elasticity Modulus and Tensile Strength are 80% infill density, 90� orientation, and PLA material type. These results not only identify the best possible configurations, but also offer a detailed insight into the complex relationship between printing settings and the mechanical qualities that are produced. A Validation tests is carried out to demonstrate that the model is reliable and has a predicted accuracy of more than 90%. This strong validation demonstrates how well the model works in real-world scenarios and gives rise to confidence in its capacity to direct additive manufacturing procedures toward improved mechanical performance. This work essentially illuminates the intricacies of 3D printing parameters and provides a clear path for maximizing mechanical qualities, making it a lighthouse in the field of additive manufacturing. The research findings offer significant assistance for engineers and designers who aim to optimize their processes and improve the mechanical properties of 3D-printed components, as industries increasingly depend on additive manufacturing for a variety of purposes. � ASM International 2024.