An efficient CFD-DEM approach for full-scale aero-compressor blade sand-erosion prediction

The sand ingestion process in aero-engines is inherently multiscale, multiphase, and highly unsteady, posing substantial challenges for the efficient and high-fidelity simulation of sand erosion on compressor blades. Conventional simulation methods require real-time resolution of transient flow fields, including rotor motion, resulting in prohibitively high computational costs and limited practicality for engineering applications. To address this long-standing efficiency bottleneck, this study proposes a novel frozen-flow-field-based method for efficient prediction of blade erosion. Based on the fundamental premise that erosion is primarily governed by the relative motion between sand particles and blades, and by exploiting the inherent circumferential periodicity of axial compressors, the proposed method incorporates rotor rotation effects by transforming particle motion relative to the blades within a stationary frozen flow field. This method eliminates the need to track absolute particle trajectories and avoids repetitive transient flow-field solutions, thereby significantly reducing computational cost while maintaining predictive accuracy. Validation against a single-stage fan sand ingestion experiment shows excellent agreement with measured data and demonstrates a computational speed-up exceeding 45 × compared with conventional transient methods. The method is further applied to simulate sand erosion in a full-scale 3.5-stage turbofan fan, successfully addressing long-standing questions concerning the effects of particle shape and size on erosion, as well as the occurrence of multiple particle-blade impacts. Overall, this work provides an efficient, accurate, and practical numerical framework for high-fidelity simulation of sand erosion in full-scale aero-engine compressors.