Predicting Layer Roughness with Weaving Path in Robotic Wire Arc Additive Manufacturing Using Multilayer Perceptron

Wire Arc Additive Manufacturing (WAAM) is an additive manufacturing technique that uses an electric arc to melt metal and deposit it layer upon layer, thus forming a 3D part. Most researchers considered weld bead width, height, and penetration as the characteristic performance. If the roughness of a deposited layer is reduced, less machining is required and material wastage is decreased. Hence, the deposition of weld beads with minimized roughness demands great attention. A few researchers who tried to minimize roughness used the straight path for material deposition. This paper proposes a method to minimize the surface roughness by optimizing welding parameters using weaving path. Moreover, various methods like Multilayer Perceptron (MLP), 2nd order regression, Design of Experiments (DOE), etc. have been used by researchers in WAAM to model the welding process for producing the desired width and height of the weld beads, and thus reducing the amount of machining. Since the welding is a highly nonlinear process, Multilayer Perceptron (MLP), also known as Artificial Neural Network (ANN) approach is used, which is an effective solution for this kind of applications, to predict the bead roughness for a given set of welding parameters- travel speed, wire feed speed, weaving amplitude, and weaving wavelength. The experimental results demonstrate that layers with minimized surface roughness have been produced.