Yifei GuanPostdoctoral Scholar

Research Focus:

Turbulence Modeling, Machine Learning, Computational Fluid Dynamics

Email:

yifeig@uchicago.edu

Biography

Preferred Pronouns: His/Him

I am currently a postdoc scholar, working in the multidisciplinary area of deep learning and fluid turbulence modeling in the Environmental Fluid Dynamics Group under the supervision of Prof. Pedram Hassanzadeh. I obtained my Ph.D. in Mechanical Engineering from the University of Washington in June 2019. I have developed expertise in computational fluid dynamics, chaotic and turbulent multi-physical fluid flow modeling, high-performance computing, and deep-learning-assisted multiscale simulations.

Research Interests

My research aims to solve grand challenges in computational simulations and finds a wide range of applications from large-scale geophysical circulation to micro-scale electro-thermo-convection. My postdoc research focuses on developing deep-learning-based data-driven multiscale models for large-eddy simulations. Specifically, I leverage deep learning to (1) discover the unclosed sub-grid terms using coarse-grained state variables, (2) ensure the stability of the online models when the discovered sub-grid terms are coupled to the numerical solver for large-eddy simulations by energy transfer analysis, and (3) generalize the data-driven sub-grid model to very different flow scenarios with higher Reynolds numbers by a transfer learning method.

Selected Publications

Guan, Y., Subel, A., Chattopadhyay, A. and Hassanzadeh, P., Learning physics-constrained subgrid-scale

closures in the small-data regime for stable and accurate LES. Physica D: Nonlinear Phenomena 2023.

Subel, A., Guan, Y., Chattopadhyay, A. and Hassanzadeh, P., Explaining the physics of transfer learning a

data-driven subgrid-scale closure to a different turbulent flow. PNAS Nexus 2023.

Guan, Y., Chattopadhyay, A., Subel, A. and Hassanzadeh, P., Stable a posteriori LES of 2D turbulence using

convolutional neural networks: Backscattering analysis and generalization to higher Re via transfer learning.

Journal of Computational Physics 2022.

Guan, Y. and Novosselov, I., Two relaxation time lattice Boltzmann method coupled to fast Fourier transform Poisson Solver: Application to Electroconvective Flow. Journal of Computational Physics 2019.

Guan, Y., Vaddi, R.S., Aliseda, A. and Novosselov, I., Experimental and numerical investigation of electrohydrodynamic flow in a point-to-ring corona discharge. Physical Review Fluids 2018.