MS087 - Stable FE Methods for Challenging Problems in Engineering and Science
Keywords: Adaptivity, Computational Methods, Data-driven methods, Stability
For a large class of partial differential equations (PDEs) arising in engineering science, e.g., singular perturbation, Helmholtz, nearly incompressible elastostatics problems, and coupled problems, the corresponding finite element (FE) discretizations suffer from a loss of stability. Certain FE discretizations
provide relief from the loss of stability by constructing FE approximations of the corresponding weak formulation to the PDE that satisfies the discrete inf-sup condition. Furthermore, recent work in data-driven methods can also be incorporated to develop stable FE schemes. To create discussions around
these issues, we invite contributions with a focus on the following:
• Stable discretization schemes for stationary and transient linear and non-linear problems,
• Residual minimization techniques such as the least squares FE method and the discontinuous Petrov-
Galerkin method and their analysis.
• A posteriori error analyses and estimates for stable discretization schemes leading to error indicators and adaptive mesh refinement strategies.
• Development of new FE basis functions leading to stable schemes.
• Implementational aspects and issues surrounding stable FE methods in modeling physical phenomena.
• Development of solvers for the linear system of equations resulting from stable discretization schemes.
• Application of stable FE methods to large-scale, complex problems in engineering science.
• Integration of data-driven techniques with FE technology.