Fast prediction of film cooling effectiveness on turbine blades under radiation effects based on regional distribution characteristics

As a key power component, turbine blades are widely used in turbomachinery such as aircraft engines. In high-temperature gas environments, designing an effective cooling scheme is essential to compensate for the insufficient heat resistance of blade materials and directly determines blade performance limits. Among various cooling methods, film cooling plays a critical role due to its high efficiency. For the rapid design of turbine blade cooling, efficient acquisition of film cooling effectiveness is indispensable and represents a key issue to be resolved. In high-temperature environments, radiation significantly affects film cooling, yet the complex nature of radiative heat transfer greatly reduces the efficiency of cooling analysis. The influence of radiation on film cooling effectiveness has not been fully explored, limiting further refinement in blade design. To address these issues, this study employed a neural network to capture the effects of key operating parameters and film cooling configurations on cooling effectiveness under radiative conditions, enabling rapid evaluation of blade film cooling. The proposed prediction network achieves high accuracy after testing. Furthermore, by incorporating Regional Blocks tailored to the spatial distribution characteristics of film cooling effectiveness, the network demonstrates an enhanced ability to characterize the cooling distribution, leading to improved prediction accuracy. The applicability of this network in turbine blade cooling design was also tested. Results show that, compared with traditional CFD methods, the established network achieves significantly higher efficiency in cooling performance evaluation. The findings of this study can support the intelligent design and radiation-related research of turbine blades, facilitating both rapid and refined blade development.