Machine learning methods have been shown to demonstrate the ability to reconstruct single-phase turbulent fluid flow from low-resolution inputs, with potential applications in many industries, but especially in engineering design. However, no work thus far has explored the application of machine learning image super-resolution methods to multiphase fluid flow. In this work, we apply the Super-Resolution Generative Adversarial Network (SRGAN) model to a multiphase turbulent fluid flow problem, specifically to reconstruct fluid phase fraction at a higher resolution. Two models were created in this work, one with a simple physics-constrained loss function and one without, and the results are discussed and analyzed. We found that both models were able to significantly outperform non-machine learning upsampling methods and can preserve an impressive amount of detail and nuance, showing the versatility of the SRGAN model for upsampling fluid simulations. But, the difference in accuracy between the two models is quite minimal, whereas physics-informed models have shown better results than non-physics-informed models in past work with single-phase fluid flow. This result leads to some important points of discussion and room for future research on the topic.
