Physics-Informed Neural Networks for Bingham Fluid Flow Simulation Coupled with an Augmented Lagrange Method

As a class of non-Newtonian fluids with yield stresses, Bingham fluids possess both solid and liquid phases separated by implicitly defined non-physical yield surfaces, which makes the standard numerical discretization challenging. The variational reformulation established by Duvaut and Lions, coupl...

Full description

Bibliographic Details

Published in:	AppliedMath
Main Author:	Jianying Zhang
Format:	Article
Language:	English
Published:	MDPI AG 2023-06-01
Subjects:	physics-informed neural network augmented Lagrange method Bingham fluid Adam algorithm Gauss–Newton method variational form
Online Access:	https://www.mdpi.com/2673-9909/3/3/28