Development of an Expert System for Prediction of Deposition Efficiency in Plasma Spraying

Deposition efficiency (DE) serves as a key performance indicator in plasma spraying, which is tailored by dozens of intrinsic and extrinsic influencing factors. Due to the nonlinear and complex interdependencies of the influencing factors, increasing DE has always been a challenging undertaking in the process development of plasma spraying. Hence, employing modern computer-aided algorithms is inevitable to overcome these complexities. In this study, an expert system is developed to predict DE from process parameters using adaptive neuro-fuzzy inference system (ANFIS) and support-vector machine (SVM). The developed expert system consists of two subsystems: (1) SVM-models from a previous work of the authors are used to predict the in-flight particle properties from different process parameters based on simulation data sets and (2) an ANFIS is developed to predict DE from in-flight particle properties based on experimental data sets. The results show that the developed expert system is able to estimate DE precisely with root-mean-square error (RMSE) of about 1.1%. The proposed system enables sustainable and cost-effective coatings through the prediction of DE for each set of process parameters.

Link to publication

Originally published at Journal of Thermal Spray Technology (Volume 32, pages 643–656, 2023)
By Robert Vaßen, Tobias Kalfhaus, Christoph Vorkötter, Yoo Jung Sohn, Susan Conze，Lutz-Michael Berger