Dynamic Icing Similarity Transformation Method Based on Dimensionless Number Sensitivity

To address accuracy challenges in scaled aircraft engine icing wind-tunnel tests caused by scale effects and multiphysics coupling, a dynamic icing similarity transformation method based on dimensionless number sensitivity is proposed. An integrated similarity analysis framework is established to systematically derive the key dimensionless parameters governing the icing process. A quantitative evaluation framework is then adopted to relate parameter variations to an icing similarity index. Furthermore, a dynamic adaptive algorithm driven by sensitivity ranking is introduced to automatically adjust similarity equation sets according to specific icing conditions, enabling coordinated multiparameter transformation. Numerical validations using a standard airfoil demonstrate that the proposed method exhibits enhanced fidelity compared to established traditional scaling methods. The results indicate that for the five selected calculation cases, the proposed method achieves higher accuracy in ice shape geometric reproduction; specifically, the icing shape error is reduced by approximately 58.26% and 46.98% compared with traditional criteria, respectively. Additionally, deviations in droplet collection efficiency distributions are controlled within 5% relative to the reference condition. By enhancing simulation accuracy and parameter adaptability, this method offers a robust and high-precision approach to similarity transformation for scaled icing tests under complex meteorological conditions.