August 17, 2025

C(NN)FD - Deep Learning Modelling of Multi-Stage Axial Compressors Aerodynamics

View Full Paper

Ask AI

Bookmark

Authors

Discussion

This framework predicts flow field and performance in multi-stage compressors, highlighting advantages over traditional CFD methods.

The framework predicts aerodynamic performance accurately, comparable to computational fluid dynamics benchmarks.
It employs a physics-based approach for dimensionality reduction, enhancing flow-field predictions across large-scale domains.
An iterative architecture improves prediction accuracy, providing quantifiable uncertainty without increasing computational cost.
The model is trained on diverse datasets including various geometries and operating conditions, ensuring generalizability across turbomachinery.

Bruni et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

August 17, 2025

View Full Paper

Ask AI

Bookmark

Authors

This framework predicts flow field and performance in multi-stage compressors, highlighting advantages over traditional CFD methods.

The framework predicts aerodynamic performance accurately, comparable to computational fluid dynamics benchmarks.
It employs a physics-based approach for dimensionality reduction, enhancing flow-field predictions across large-scale domains.
An iterative architecture improves prediction accuracy, providing quantifiable uncertainty without increasing computational cost.
The model is trained on diverse datasets including various geometries and operating conditions, ensuring generalizability across turbomachinery.

Bruni et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark