Physics Informed Neural Network-based Computational Method for Accelerating Time-Periodic Unsteady CFD Simulations

TL;DR

This paper introduces a meshless Physics Informed Neural Network (PINN) method to directly compute time-periodic solutions in CFD, significantly reducing computational time compared to traditional transient approaches.

Contribution

The paper proposes a novel PINN-based periodic CFD solver that directly targets the periodic state, bypassing initial transients, and demonstrates its efficiency for 2D heat diffusion and fluid flow problems.

Findings

PINN-based solver reduces computational time substantially.

Achieves similar accuracy to traditional methods.

Effect of hyperparameters on performance is analyzed.

Abstract

Presently, there is a steady state approach in Computational fluid dynamics (CFD) to obtain a steady solution directly from the steady state governing equations. Whereas, for obtaining a time-periodic flow solution, the present unsteady governing equations-based CFD approach starts from an initial condition and requires a large computational time during the initial non-periodic transient phase before reaching the periodic state. For obtaining the periodic flow directly, without transient simulations that may not be of interest, our objective is to propose a Physics Informed Neural Network (PINN)-based periodic CFD approach. The motivation is a substantial reduction in computational time by a meshless PINN-based periodic CFD solver as compared to the present mesh-based transient-to-periodic solver. Proof-of-concept, for the periodic CFD approach, is demonstrated here for 2D periodic heat diffusion and fluid flow problems. The proposed PINN-based periodic solver primarily focuses on the time-periodic state, optimizing the neural network model's trainable parameters to precisely fit a smaller time window (one time-period) rather than the temporal domain starting from the initial condition. After presenting a verification study, effect of the PINN-related various hyperparameters such as the number of collocation points, neural network architecture, and point spacing for numerical differentiation, on computational time and accuracy are presented. Our results demonstrate that the PINN-based periodic solver takes substantially less computational time to achieve almost same accuracy as that obtained by the traditional transient-to-periodic solver.