Towards Inline Property Prediction in Material Extrusion-Fabricated Structures with Limited Information

Modeling polymer material extrusion additive manufacturing (AM) for prediction of mechanical properties requires substantial training data, and many modeling approaches are computationally intensive. This work investigates measurement and modeling approaches to enable inline property prediction in filament-fed material extrusion AM, commonly referred to as fused filament fabrication (FFF). We explore what kinds and how much information is necessary to create an acceptably accurate model, and to output accurate predictions. We also performed detailed analysis of the types of fracture that occurred across a range of processing conditions and relate this mechanical behavior to print conditions. This work contributes to efforts towards higher reliability, higher performance additively manufactured parts.