Deep learning modelling of manufacturing and build variations on multi-stage axial compressors aerodynamics

arXiv:2310.04264v3 Announce Type: replace
Abstract: Application of deep learning methods to physical simulations such as CFD (Computational Fluid Dynamics) for turbomachinery applications, have been so far of limited industrial relevance. This paper demonstrates the development and application of a deep learning framework for real-time predictions of the impact of manufacturing and build variations, such as tip clearance and surface roughness, on the flow field and aerodynamic performance of multi-stage axial compressors in gas turbines. The associated scatter in compressor efficiency …

abstract aerodynamics application applications arxiv build cfd computational cs.ce cs.lg deep learning deep learning framework development dynamics fluid dynamics framework industrial manufacturing modelling paper physics.flu-dyn predictions real-time simulations stage type