Prediction of mechanical properties of short fiber reinforced composite fabricated by Fused Filament Fabrication (FFF) method using Machine Learning
The tremendous increase in the application of additive manufacturing (AM) has gained much attention in recent times due to its usability and capacity to ascribe improved mechanical properties on printed parts with no tooling. AM process with the use of FFF is becoming an integral fabrication method for producing the complex geometries and machine components with intricate parts. There is a corresponding increase in dataset derived from AM process which has ushered the use of highly computational models like machine learning (ML) and deep learning for analysis, prediction, classification, dimensional accuracy, and optimization of methods and printing properties of fabricated parts.
This study explores the contribution of printing parameters, e.g., printing speed, layer height and infill density on mechanical properties of short carbon fiber samples produced using FFF technology. ML models will be used for classification of samples built with different print parameters, the models will analyze microstructural images captured under microscope as input dataset and make prediction and classification based on their microstructural attributes (bead shape). In this study, the computation ability of ML models will be used in the predictions for improved mechanical properties based off results of tensile tests conducted on FFF material samples with various printing parameters. The findings of this study provide evidence and insight that ML can be used to optimize printing performance and its applications.