Experimental and Investigation of ABS Filament Process Variables on Tensile Strength Using an Artificial Neural Network and Regression Model

Abstract

Fused deposition modeling (FDM) is a commonly used 3D printing technique that involves heating, extruding, and depositing thermoplastic polymer filaments. The quality of FDM components is greatly influenced by the chosen processing settings. In this study, the Taguchi technique and artificial neural network were employed to predict the ultimate tensile strength of FDM components and establish a mathematical model. The mechanical properties of ABS were analyzed by varying parameters such as layer thickness, printing speed, direction angle, number of parameters, and nozzle temperature at five different levels. FDM 3D printers were used to fabricate samples for testing, following the ASTM-D638 standards, using the Taguchi orthogonal array experimental design method to set the process parameters. The results indicated that the printing process factors had a significant impact on tensile strength, with test values ranging from 31 to 38 MPa. The neural network achieved a maximum error of 5.518% when predicting tensile strength values, while the analytical model exhibited an error of 19.376%.