Prediction of tensile strength in fused deposition modeling process using artificial neural network and fuzzy logic

Abstract

Fused deposition modeling is a modern rapid prototyping technique that is used for swiftly replicating concept modeling, physical modeling, and end-of-line manufacture. Precision parameter selection is crucial for generating high-quality products with excellent mechanical properties, such as tensile strength. This study looked at three essential process variables: infill density, extruder temperature, and print speed. The relationship between these parameters and tensile strength of printed polylactic acid components was investigated. Artificial neural network (ANN) and Fuzzy logic (FL) method are utilized to develop a prediction model. The test samples have been printed using a 3D forge Dreamer II FDM printing machine. In Minitab software, the response surface design of the Box–Behnken technique with 15 experimental sets was used to organize the trials. The results revealed that extruder temperature and print speed had a minor impact on tensile strength; however, infill density has a large impact. The ANN and FL models all predicted tensile strength with a high degree of accuracy, with maximum absolute percentage errors of 2.21%, and 3.29%, respectively. The model and the experimental data were found to be in good agreement, according to the findings. Furthermore, when compared to FL modeling, ANN models with arithmetical value indices were the best predictive model.