Numerical approach to estimate the elastic mechanical properties of FFF printed parts according to the infill density

Abstract

Additive Manufacturing (AM) processes built-up parts by adding material layer by layer. Fused Filament Fabrication (FFF) is one of the most popular AM plastic technologies used in various fields such as biomedical, automotive, aerospace or prototyping engineering, among others. One of the main characteristics of FFF parts is their non-homogeneous behavior as a result of the internal structural geometry, commonly known as the infill pattern. As a result, printed parts typically exhibit a macroscopic orthotropic structural behavior. The Finite Element Method (FEM) can be used to predict the macroscopic elastic constitutive behavior of FFF printed parts through several Finite Element analyses (FEA) of a Representative Volume Element (RVE). The elastic constants can be calculated by means of a homogenization procedure. In this paper, several finite element analyses of RVE models are carried out in order to train and test Machine Learning (ML) models to estimate the orthotropic longitudinal and transverse elastic parameters of FFF printed parts according to the infill density for a 0°/90° linear infill pattern. Linear Regression (LR) and Artificial Neural Network (ANN) ML models are successfully trained and tested by means of data obtained from RVE numerical models. Finally, ML predictions are compared with experimental results, showing differences lower than 10% in all analyzed cases.