Prediction of film-cooling effectiveness based on support vector machine

Least square support vector machine (LS-SVM) model is applied to predict the lateral averaged adiabatic film-cooling effectiveness on a flat plate surface downstream of a row of cylindrical holes. The dataset used to develop and validate the presented model is obtained from the public literature. The input parameters of LS-SVM include dimensionless downstream distance, pitch-to-diameter ratio, hole incline angle, hole compound angle, length-to-diameter ratio, blowing ratio, density ratio, and mainstream turbulence intensity. The predicted results are found to be in good agreement with the experimental results (the mean relative error is about 17.5%). The comparison between LS-SVM model and existing semi-empirical correlations is carried out, and the prediction performance of LS-SVM model is much better. Moreover, the effects of LS-SVM input parameters on film-cooling effectiveness are discussed in detail. LS-SVM is a promising model to predict the film-cooling effectiveness.