Aerodynamic shape design is essential for improving aircraft performance and efficiency. First, this study introduces a data-driven optimization framework utilizing a multi-fidelity convolutional neural network (MFCNN) for aerodynamic shape optimization. To achieve better optimization results with reduced computational cost, the framework dynamically incorporates new data in each optimization cycle. Specifically, it constantly involves the optimal solution from previous cycle as a new high-fidelity sample and employs a low-fidelity infilling strategy that maximizes the minimum Euclidean distance for selecting new low-fidelity samples. Moreover, a standard synthetic benchmark is used to elaborate the procedure of optimization and show the capability and effectiveness of the framework. Finally, the framework is applied to two aerodynamic shape optimization problems: maximizing the lift-to-drag ratio for the Royal Aircraft Establishment 2822 (RAE2822) airfoils and minimizing the cruise drag coefficient for the three-dimensional (3D) drooped and scarfed non-axisymmetric nacelles. The framework increases the lift-to-drag ratio by 51.21% over the baseline and achieves an 18.79% reduction in the cruise drag coefficient for nacelle optimization, outperforming traditional multi-fidelity deep neural network optimization framework. Sufficiently utilizing the implicit relations between different fidelity levels of data through defined local perceptual fields and convolution, our MFCNN-based optimization framework signifies a step forward in the efficiency and accuracy of aerodynamic shape optimization.
Leakage flow through shrouded-stator cavities introduces additional aerodynamic losses and near-hub flow distortions, which may propagate through downstream blade rows and deteriorate compressor performance. To enable computationally efficient prediction of these leakage-induced effects, this study develops a reduced-order leakage prediction framework that avoids the explicit construction and meshing of internal cavity geometries. The proposed coupling framework accounts for the thermodynamic and aerodynamic characteristics of the leakage flow by incorporating corrections for the total-pressure loss and windage heating within the seal cavity. The resulting leakage state is then transferred to the mainstream computational domain through physically consistent hub-wall mapping, thereby enabling effective coupling with the three-dimensional mainstream flow solver. The model is first validated against representative seal configurations and a high-load compressor cascade, and is further applied to realistic single-stage and 3.5-stage compressor. The results demonstrate that the proposed model accurately reproduces the clearance-dependent leakage characteristics, near-hub total pressure deficit, and flow angle distortion, while maintaining good agreement with the overall compressor performance trends. For the investigated compressor configurations, the predicted overall performance parameters exhibit differences of less than 1% compared with those obtained using fully resolved three-dimensional simulations, while the computational mesh size is reduced by approximately 30%. Furthermore, the model effectively captures the generation of leakage-induced near-hub flow disturbances, their spatial diffusion, and their interstage propagation through successive blade rows. These results demonstrate that the proposed framework provides a physically consistent and computationally efficient approach for predicting leakage-induced losses in multistage compressors equipped with shrouded stator cavities.