Moisture absorption study and mechanical property prediction on 3D printed parts using hybrid neural network models

The effect of printing layer thickness and infill percentage on the water intake ability and mechanical properties of 3D printed PLA (Poly Lactic Acid) parts are explored in this research article. This article is of particular interest because PLA components are typically hydrophilic, preventing their application in humid environments. Tensile and bending test samples were prepared using 1.75 mm PLA filament with infill percentages varying from 50-100% and 0.1-0.3 mm layer thickness. The design of experiments was made using a central composite design to perform the experiments. The water absorption tests were conducted as per ISO 62:2008. The bending and tensile properties increases with the increase in infill percentage and minimum layer thickness. Adaptive Neuro-Fuzzy Interface System (ANFIS) and Artificial Neural Network-Teaching Learning Based Optimization Algorithm (ANN-TLBO) was used to develop an adaptive model between the input parameters and the mechanical properties. The models were evaluated using statistical indices such as coefficient of determination (R2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R<^>2$$\end{document}), root mean square error (RMSE) and mean absolute error (MAE). The R2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R<^>2$$\end{document}, RMSE and MAE values obtained using ANN-TLBO for tensile and bending tests were 0.99, 0.01 and 0.99, 0.008 and 0.11, 0.12 respectively. It is observed that ANN-TLBO predicted close results with the experimental value for tensile and bending strength.

Automated summary

This research article explores the impact of printing layer thickness and infill percentage on the water intake ability and mechanical properties of 3D printed PLA parts. The results show that increasing infill percentage and reducing layer thickness can improve the bending and tensile properties of the parts. Additionally, an adaptive model between the input parameters and mechanical properties is developed using ANFIS and ANN-TLBO methods.