Neural network application to FDM surface roughness prediction

Fused deposition modeling (FDM) is the second most spread Additive Fabrication (AF) technique in the world. It fabricates, layer by layer, objects by depositing acrylonitrile butadiene styrene (ABS) in filament form. No complete and reliable knowledge about the surface quality achievable, lead to manufacturing subsidiaries, or to replace the AF with subtractive techniques with significant added costs. So, roughness prediction models are necessary in order to save time and costs. In this work a roughness parameter models reliable has been developed by using Artificial Neural Network (ANN).