Speaker
Description
Numerical simulations of turbulent flows represent a significant challenge in fluid dynamics due to their complexity and high computational cost. High-resolution techniques such as Large Eddy Simulation (LES) are able to capture fine details of turbulent structures but are generally not computationally affordable, particularly for technologically relevant problems. Recent advances in machine learning, specifically in generative probabilistic models, offer promising alternatives for turbulence modeling to overcome this issue. For example, our previous work introduced generative adversarial networks (GAN) as a mathematically well-founded approach for turbulence modeling and demonstrated the generalization capabilities of GAN-based synthetic turbulence generators under geometric flow configuration changes. In this work, we compare three generative models - Variational Autoencoders (VAE), Deep Convolutional GAN (DCGAN), and Denoising Diffusion Probabilistic Models (DDPM) - in simulating a 2D Kármán vortex street around a fixed cylinder. Training data was obtained by means of LES. We evaluate each model's ability to capture the statistical properties and spatial structures of the turbulent flow. Our results demonstrate that DDPM and DCGAN effectively replicate the flow distribution, highlighting their potential as efficient and accurate tools for turbulence modeling. We find a strong argument for DCGAN, as although they are more difficult to train, they gave the fastest inference and training time, require less data to train compared to VAE and DDPM, and provide the results most closely aligned with the input stream. In contrast, VAE train quickly and can generate samples quickly but does not produce adequate results, and DDPM, whilst effective, is significantly slower at both inference and training time.
