| This work presents an analytical model of the temperature/concentration dynamic distributions and microstructure growth that occur during the production of intermetallic (e.g., nickel aluminide) coatings on a variety of substrates, as for example steel. The temperature evolution, component dissolution, and intermetallic compound growth arise by reactive thermal processing of layered elemental precursors using a moving heat source of Gaussian power distribution, such as a plasma arc, that scans the surface of the coating. The model, which is computationally parallel and meshless (i.e., decoupled with the capability to be solved numerically in real time), is based on kinetic growth theories, lumped energy and mass balances, and convolution expressions of distributed temperature and concentration Green's fields (accounting for the orientation of their gradient and decomposing heat and mass transfer across the coating from substrate conduction). The numerical results are validated with Ni-Al coatings processed on a robotic plasma arc laboratory station, through in-process infrared thermal sensing and off-line metallographic analysis. It is shown that the predicted coating temperature, dissolution penetration value, and thickness of the coating layer compare well with the experimentally obtained results. This suggests the applicability of the model to address the growth of intermetallics in melts of uniform and non-uniform initial solute concentration, and supports its use as a real time basis for design and/or adaptation of a thermal control system for the coating process. Based on these results, a distributed parameter control method, involving on-line parameter identification and model adaptation, is developed using the model for in-process estimation and control of the coating thickness achieved during the thermal/reaction process. |
