Investigations and predictions for mechanical and surface properties of FFF prints using DOE, ML and FEA

In the past two decades, number of studies have been reported on process parametric optimisation of the fused filament fabrication (FFF) process for polylactic acid-based functional prototypes by using design of experiment techniques. But hitherto little has been reported on the prediction of mechanical and surface properties of FFF prints using machine learning (ML) and a finite element analysis (FEA) approach. The present research work aims to provide insight into the influence of FFF parameters, namely: layer thickness (LT), infill structure (IS) and infill density (ID) on Young's modulus and surface roughness. In the first stage, optimum process parameters of FFF for Young's modulus and surface roughness have been evaluated as 0.16 mm LT, cubic IS and 45% ID along with an analysis of variance-based linear models. The ML-based computational model has been developed to predict and verify the experimental observations in the second stage. Finally, in the third stage, experimental and predicted values of Young's modulus have been co-related based upon FEA (by using ABAQUS software).

Automated summary

In the past two decades, numerous studies have focused on process parametric optimisation of fused filament fabrication (FFF) for polylactic acid-based functional prototypes using design of experiment techniques. However, there is limited research on predicting mechanical and surface properties of FFF prints through machine learning (ML) and finite element analysis (FEA) methods. This study aims to investigate the impact of FFF parameters, such as layer thickness, infill structure, and infill density, on Young's modulus and surface roughness. The research involves evaluating optimal process parameters, developing an ML-based computational model for prediction, and correlating experimental and predicted values using FEA.