Mechanical Properties of 3D-Printed Components Using Fused Deposition Modeling: Optimization Using the Desirability Approach and Machine Learning Regressor

Abstract

The fused deposition modelling (FDM) technique involves the deposition of a fused layer of material according to the geometry designed in the software. Several parameters affect the quality of parts produced by FDM. This paper investigates the effect of FDM printing process parameters on tensile strength, impact strength, and flexural strength. The effects of process parameters such as printing speed, layer thickness, extrusion temperature, and infill percentage are studied. Polyactic acid (PLA) was used as a filament material for printing test specimens. The experimental layout is designed according to response surface methodology (RSM) and responses are collected. Specimens are prepared for testing of these parameters as per ASTM standards. A mathematical model for each of the responses is developed based on the nonlinear regression method. The desirability approach, nonlinear regression, as well as experimental values are in close agreement with each other. The desirability approach predicted the tensile strength, impact strength, and flexural strength with a less percentage error of 3.109, 6.532, and 3.712, respectively. The nonlinear regression approach predicted the tensile strength, impact strength, and flexural strength with a less percentage error of 2.977, 6.532, and 3.474, respectively. The desirability concept and nonlinear regression approach resulted in the best mechanical property of the FDM-printed part.

Optimization of tensile strength in 3D printed PLA parts via meta-heuristic approaches: a comparative study

“…This process involves preparing a computer-aided FIGURE 1 (A) FDM printing setup (Potnis et al, 2023). (B) Schematic sketch of Tensile Specimen (Jatti et al, 2022). (C) 3D printed tensile specimen (Potnis et al, 2023).…”

Prediction of Optimal Parameter Settings and Significant Parameter for Reduced Geometric Deviations Through Multi Criteria Decision Making and Machine Learning Algorithms

“…The adoption of ML algorithms in FDM parameter optimization has been witnessed recently. Vijaykumar et.al [21] optimized FDM process parameters using desirability approach and machine learning regressor for PLA material. The models developed by both the methods are able to predict tensile, impact and flexural strength with low error percentage.…”

Using Bayesian Regularized Artificial Neural Networks to Predict the Tensile Strength of Additively Manufactured Polylactic Acid Parts

“…The same printing processes were investigated by Jatti et al using an ML nonlinear regression algorithm only for tensile strength prediction. The results were able to predict the tensile strength with a percentage error of less than 2.977 [32]. Models for predicting the ultimate tensile strength were developed using an ANN.…”