Prediction of Flatness and Deposition Efficiency in Aluminum Wire Explosion Spraying Based on Small Sample Data

Metal wire explosive spraying, a novel method for green coating preparation, exhibits an exceptionally high jetting velocity, facilitating metallurgical bonding between the coating and the substrate. The bonding strength achieved is significantly higher than that of traditional plasma spraying and flame spraying methods. Ensuring the flatness of the coatings and improving the coating deposition efficiency are key factors for the industrial application of this technique. To predict coating flatness and deposition efficiency, an experimental setup for aluminum wire explosive spraying was constructed. Through a mechanistic analysis, four process parameters—capacitance, initial charging voltage, wire diameter, and spraying distance—were selected to investigate their impact on aluminum coating roughness and deposition efficiency. To address the issue of small sample sizes due to variability in equipment parameters and environmental conditions, a Genetic Algorithm-Random Forest (GA-RF) model was developed to predict coating roughness and deposition efficiency based on small sample data. The results show that the model performs excellently in terms of evaluation metrics and can achieve accurate predictions, laying a foundation for the optimization of process parameters in aluminum coating preparation via explosive spraying.