The complex flow modeling based on machine learning is becoming a promising way to describe multiphase fluid systems. This work demonstrates how a physics-informed neural network promotes the combination of traditional governing equations and advanced interface evolution equations without intricate algorithms. We develop physics-informed neural networks for phase-field method (PF-PINNs) in two-dimensional immiscible incompressibletwo-phase flow. The Cahn-Hillard equation and Navier-Stokes equationsare encoded directly into the residuals of a fully connected neural network. Compared withthe traditional interface-capturing method, the phase-field model has a firm physical basis because it is based on the Ginzburg-Landau theory and conserves mass and energy. It also performs well in two-phase flow at large density ratio. However, the high-order differential nonlinear term of the Cahn-Hilliard equation poses a great challenge for obtaining numerical solutions. Thus, in this work, we adopt neural networks to tackle the challenge by solving high-order derivate terms and capture the interface adaptively. To enhance the accuracy and efficiency of PF-PINNs, we use the time-marching strategy and the forced constraint of the density and viscosity. The PF-PINNs are tested by two cases for presenting the interface-capturing ability of PINNs and evaluating the accuracy of PF-PINNs at large density ratio (up to 1000). The shape of the interface in both cases coincides well with the reference results, and the dynamic behavior of the second case is precisely captured.We also quantify the variations in the center of mass and increasing velocity over time for validation purposes. The results show that PF-PINNs exploit the automatic differentiation without sacrificing the high accuracy of the phase-field method.

Abstract. The prediction of a wind farm near the wind turbines has a significant effect on the safety as well as economy of wind power generation. To assess the wind resource distribution within a complex terrain, a computational fluid dynamics (CFD) based wind farm forecast microscale model is developed. The model uses the Reynolds Averaged Navier-Stokes (RANS) model to characterize the turbulence. By using the results of Weather Research and Forecasting (WRF) mesoscale weather forecast model as the input of the CFD model, a coupled model of CFD-WRF is established. A special method is used for the treatment of the information interchange on the lateral boundary between two models. This established coupled model is applied in predicting the wind farm near a wind turbine in Hong Gang-zi, Jilin, China. The results from this simulation are compared to real measured data. On this basis, the accuracy and efficiency of turbulence characterization schemes are discussed. It indicates that this coupling system is easy to implement and can make these two separate models work in parallel. The CFD model coupled with WRF has the advantage of high accuracy and fast speed, which makes it valid for the wind power generation.

No abstract