A correction method based on CGAN for scaling criteria of turbine blades in high radiation environments

The scaling criteria can lower the required hot-gas temperature for testing turbine blade film-cooling effectiveness, thereby significantly facilitating the process. However, the current scaling criteria do not consider the influence of radiation, leading to a notable disparity between the low-temperature film-cooling effectiveness obtained from testing and the desired high-temperature film-cooling effectiveness. This study aims to extract the relationship between radiation influence and adiabatic film-cooling effectiveness by using a Conditional Generative Adversarial Network (CGAN). The radiation-influencing parameters are hot-gas temperature, temperature ratio, wall emissivity, gas radiation absorption coefficient and blackbody radiation coefficient. The incorporation of radiation effects into the scaling process will ultimately result in the correction of the scaling criteria. The findings demonstrated that the CGAN model is capable of accurately forecasting the high-temperature film-cooling effectiveness based on the low-temperature film-cooling effectiveness within the scaling process. Furthermore, the prediction accuracy is approximately 10 times higher than that of conventional scaling criteria, which markedly minimizes the prediction error pertaining to the high-temperature data due to the exclusion of radiation influence in the conventional scaling criteria. The findings of this research have the potential to eliminate the need for extensive high-temperature testing, thus reducing energy consumption and pollutant emissions.