In the production of thermal barrier coating (TBC) with the atmospheric plasma spray coating system, the process parameters directly affect the production cost and performance of the coatings. In this study, a comprehensive modeling-design-optimization study was conducted to improve the analytical performance of TBC. For this purpose, the data were taken from a literature study that included an extensive experimental design application. The modeling study prepared first, second, and third-order polynomial, trigonometric, and logarithmic-based models for each process output. Model selections were made with neuro-regression and a statistical method. The selected models were run on four different stochastic optimization algorithms for the coatings' deposition efficiency, bond strength, porosity, and hardness value outputs. Thirty-six neuro-regression models prepared in the modeling study have high R2training values. The second-order logarithmic nonlinear (SOLN) models were successful in the coatings' deposition efficiency and bond strength, and the polynomial nonlinear models were successful for the four process outputs. Therefore, they were chosen as the objective functions of the optimization algorithms. In addition, the selected models were run at the parameters determined by numerical optimization in the reference publication, and the prediction abilities of the models in the two studies were compared. SOLN models for deposition efficiency and bond strength values, second-order nonlinear model for hardness value, and reference study’ model predicted more closely to the validation test result for porosity values of coating. In the optimization studies, three or more algorithms suggested the same results with the same parameter sets for the deposition efficiency and hardness values. The optimization results show that these points can be a global optimum point for optimizing these two coating properties.
The author thanks reviewers for constructive comments and suggestions that helped to improve the quality of the article.
Primary Language | English |
---|---|
Subjects | Engineering |
Journal Section | Research Articles |
Authors | |
Publication Date | June 30, 2022 |
Submission Date | March 9, 2022 |
Published in Issue | Year 2022 Issue: 049 |