Automated Non-Destructive Inspection of Fused Filament Fabrication Components Using Thermographic Signal Reconstruction

Abstract Manufacturers struggle to produce low-cost, robust and intricate components in small batches. Additive processes like Fused Filament Fabrication (FFF) inexpensively generate such complex geometries, but potential defects may limit these components’ viability in critical applications. We present a high-accuracy, high-throughput and low-cost approach to automated non-destructive testing (NDT) for FFF interlayer delamination. This Artificially Intelligent (AI) approach utilizes Flash Thermography (FT) data processed with Thermographic Signal Reconstruction (TSR). A Deep Neural Network (DNN) attains 95.4% per-pixel accuracy when differentiating four delamination severities 5 mm below the surface in PolyLactic Acid (PLA) widgets, and 98.6% accuracy in differentiating acceptable from unacceptable states for the same components. Automation supports time- and cost-efficient inspection for delamination defects in 100% of widgets, supporting FFF's use in critical and lot-size one applications.