Image Upsampling of Low Resolution Turbulent CFD Domains with U-Net


METADATA ONLY
Loading...

Date

2025

Publication Type

Conference Paper

ETH Bibliography

yes

Citations

Altmetric
METADATA ONLY

Data

Rights / License

Abstract

The modelling of turbulent airflow with CFD is a computationally expensive task, yet vital for assessing architectural and urban ventilation concepts. Simplified airflow prediction methods exist, however at the cost of lacking accuracy and/or precision. This paper therefore proposes a hybrid simulation and deep learning approach. We utilize images of low resolution CFD domains as input to a U-Net neural network. The generated output is a high resolution upsampled image. As training data and application case, we use turbulent indoor flow with forced convection. Results are promising, as the generated flow fields can recreate higher level of details from the low resolution inputs as when compared to bicubic interpolation. However, the approach leaves room for improvement especially with respect to generated image sharpness.

Permanent link

Publication status

published

Book title

Multiphysics and Multiscale Building Physics. Proceedings of the 9th International Building Physics Conference (IBPC 2024). Volume 4: Indoor Air Quality (IAQ), Lighting and Acoustics

Volume

555

Pages / Article No.

48 - 54

Publisher

Springer

Event

9th International Building Physics Conference (IBPC 2024)

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Deep Learning; Convolutional Neural Network; Upsampling; Computational Fluid Dynamics; Large Eddy Simulation

Organisational unit

08060 - FCL / FCL
03902 - Schlüter, Arno / Schlüter, Arno

Notes

Funding

Related publications and datasets