Prediction of Surface Roughness of SLM Built Parts after Finishing Processes Using an Artificial Neural Network

A known problem of additive manufactured parts is their poor surface quality, which influences product performance. There are different treatments to improve quality: blasting commonly employed mechanical properties and reduce roughness, electropolishing clean shot peened surfaces the roughness. However, final roughness conditioned by multiple parameters related these techniques. This paper presents a prediction model (Ra) using an Artificial Neural Network considering two SLM manufacturing process seven processes. proven be in agreement with 429 experimental results. Moreover, this then used find optimal conditions applied during order roughly 60%.