Description and Prediction of Multi-layer Profile in Cold Spray Using Artificial Neural Networks

Cold spray is a newly developed solid-state metal deposition technology, which allows for preparing various functional coatings and repairing damaged metal components, as well as fabricating freestanding parts. In order to obtain the deposits with the desired shape and accuracy, the coating profile, including its thickness and distribution, is an important indicator to monitor and control. In this work, an artificial neural network (ANN) model has been employed to describe and predict the multi-layer profile of cold-sprayed deposits. Compared to conventional feature-based modeling methods, the ANN model is capable of simulating a complete track profile on defined substrate morphologies. The superiority of the ANN approach is further emphasized by its ability to simulate a multi-layer profile, which differs from previous works that focus on single-layer profiles. It is essential for guiding the coating formation and fabrication of near-net-shape parts. The results imply that the ANN model is well trained and capable of predicting multi-layer profiles with acceptable accuracy. It can be used for profile control during cold spray additive manufacturing.