Design of Experiment Using AI for Laser Heat Treatment Process of Cold Sprayed Copper Coating

Laser surface heat treatment process is a technology of choice where specific localized heating is required. It guarantees a more precise and cost-effective alternative to other surface treatments specifically for large parts composed of dissimilar material combinations or for temperature sensitive cases such as used nuclear fuel containers. This work relies on designing experimental procedures to explore optimized laser parameters using artificial intelligence simulation. An artificial neural networks was developed and used to find the laser heat treating parameters to be applied to as-deposited copper cold spray coating to obtain a 30% through-thickness hardness reduction. Thermal histories were obtained using thermal modeling of the laser heating process and the resulting hardness reductions obtained experimentally. The ANN model was developed using local thermal history as model input and local hardness reduction as model output. Deploying thermal history as input allows to train the network and to use it for any type of heating process and any specimen dimension. It is demonstrated that the model is capable of predicting the local hardness reduction of part according to the local thermal history resulting from various laser heat treatment conditions and thus, can eventually be used to assist in process parameters selection and optimization.