Locking-free and gradient-robust \(H(\operatorname{div})\)-conforming HDG methods for linear elasticity (Q1995983)

In this paper, the concept of gradient-robustness for numerical methods for linear elasticity is introduced. A special class of discretizations for linear elasticity is considered: \(H(\operatorname{div})\)-conforming Hybrid Discontinuous Galerkin (HDG) discretizations by keeping track of the volume-locking and gradient-robustness property of the method. Two efficient variants of a divergence-conforming HDG scheme with reduced globally coupled degrees of freedom are discussed and analyzed. The importance of gradient-robustness and the computational efficiency of the proposed relaxed \(H(\operatorname{div})\)-conforming HDG method is demonstrated on a linear multi-physics thermo-elasticity problem in 2D and 3D. It was shown that the gradient-robustness against strong gradients fields in \(L^2\) leads to much more accurate schemes for nearly incompressible materials