3D printing of acrylonitrile butadiene styrene by fused deposition modeling: artificial neural network and response surface method analyses.

Abstract

Additive manufacturing of acrylonitrile butadiene styrene (ABS) was investigated based on statistical analysis via an optimization method. The present article discusses the influence of the layer thickness (LT), infill percentage (IP), and contours number (C) on the maximum failure load and elastic modulus of the final product of ABS. ABS is a low-cost manufacturing thermoplastic that can be easily fabricated, thermoformed, and machined. Chemical, stress, and creep resistance is all excellent in this thermoplastic material. ABS combines a good balance of impact, heat, chemical, and abrasion resistance with dimensional stability, tensile strength, surface hardness, rigidity, and electrical properties. To comprehend the impact of additive manufacturing parameters on the build quality, both artificial neural network (ANN) and response surface method (RSM) were used to model the data. The main characteristics of the build considered for modeling were ultimate tensile strength (UTS) and elastic modulus. Main effect plots and 3d plots were extracted from ANN and RSM models to analyze the process. The two models were compared in terms of their accuracy and capability to analyze the process. It was concluded that though ANN is more accurate in the prediction of the results, both tools can be used to model the mechanical properties of ABS formed by 3D printing. Both models yielded similar results and could effectively give the effect of each variable on the mechanical properties.