A physics-constrained neural network for multiphase flows

The present study develops a physics-constrained neural network (PCNN) to predict sequential patterns and motions of multiphase flows (MPFs), which includes strong interactions among various fluid phases. To the order parameters, locate individual phases in future time, (NN) is applied quickly infer dynamics by encoding observations. consistent conservative boundedness mapping algorithm (MCBOM) next implemented correct predicted parameters. This enforces parameters strictly satisfy mass conservation, summation volume fractions be unity, consistency reduction, Then, density mixture updated from corrected Finally, velocity time another NN with same structure, but conservation momentum included loss function shrink parameter space. proposed PCNN for MPFs sequentially performs (NN)-(MCBOM)-(NN), avoids nonphysical behaviors accelerates convergence, requires fewer data make predictions. Numerical experiments demonstrate that capable predicting effectively.