Finite element de Rham and Stokes complexes in three dimensions

Long Chen, Xuehai Huang

TL;DR

This paper provides a systematic framework for constructing finite element de Rham and Stokes complexes in three dimensions with customizable levels of smoothness ($C^m$ continuity). By using a novel geometric decomposition of simplicial lattices, the authors derive stable, exact sequences of functional spaces that are essential for developing robust numerical solvers for complex partial differential equations (PDEs) like the Stokes equations.

What Problem Does This Paper Address?

In the world of numerical simulation, simply "discretizing" a physics equation isn't enough. For problems involving fluid flow (Stokes equations) or electromagnetics (Maxwell’s equations), the discrete spaces used to represent different physical quantities (like velocity and pressure) must relate to each other in a very specific way. This relationship is formalized as a Hilbert complex—a sequence of spaces where the range of one differential operator (like the gradient) is exactly the kernel of the next (like the curl).

While these complexes are well-understood in 2D, extending them to 3D with high-order smoothness (ensuring functions are continuous not just in value, but in their derivatives across element boundaries) is notoriously difficult. Previous methods were often ad-hoc or limited to specific cases. This paper addresses the need for a unified, systematic construction of these complexes in 3D, ensuring both exactness (no "gaps" in the sequence) and div-stability (crucial for preventing numerical artifacts in fluid simulations).

Methods

The authors utilize a sophisticated mathematical toolkit centered on Simplicial Lattices and Barycentric Calculus.

  1. Geometric Decomposition: Instead of treating a tetrahedron as a single block, the authors partition the "lattice" of polynomial degrees of freedom (DoFs) based on their "distance" to sub-simplices (vertices, edges, and faces). This allows them to precisely control smoothness ($C^r$ continuity) at each geometric level.
  2. Bernstein Basis: They leverage Bernstein polynomials, which are naturally suited to barycentric coordinates, making it easier to define and enforce continuity across the shared faces of tetrahedra.
  3. Grafting Complexes: The paper introduces a "grafting" technique. They start with a standard de Rham complex and "graft" on additional smoothness requirements to create new complexes, such as the Stokes complex.
  4. Commutative Diagrams: To ensure the discrete model faithfully represents the continuous physics, they construct "interpolation operators" that satisfy commutative properties—meaning that interpolating a function and then taking its gradient is the same as taking the gradient first and then interpolating.

Datasets

As this is a theoretical contribution to numerical analysis and finite element theory, it does not use empirical datasets. The "evaluation" consists of rigorous mathematical proofs of the properties (stability, exactness, and unisolvence) of the newly constructed finite element spaces.

Key Insights and Ideas

  • The "Distance" Metric: The core "aha moment" is using the algebraic distance of a lattice node to a face or edge to determine its contribution to the function's smoothness. This turns a complex geometric continuity problem into a more manageable combinatorial one.
  • Unified Construction: Rather than treating $H(\text{div})$ and $H(\text{curl})$ elements as separate entities, the authors show they can be derived from the same scalar smooth finite element framework.
  • Stability through Bubble Spaces: They demonstrate that the "div-stability" (the LBB condition) of a velocity-pressure pair can be guaranteed by carefully constructing "bubble spaces"—polynomials that vanish on the boundaries of the tetrahedron—to fill the gaps in the divergence operator's range.

Key Results

  • New 3D Families: The paper successfully constructs 3D versions of famous 2D elements, including the Hermite and Argyris families, integrated into complete de Rham and Stokes complexes.
  • Exactness Proofs: The authors prove that their constructed sequences are exact on any domain homeomorphic to a ball, meaning they are free of the "spurious modes" that plague unstable numerical methods.
  • Generalization: While focused on 3D, the authors provide an appendix showing how this lattice decomposition generalizes to arbitrary dimensions ($n$-D), providing a roadmap for even higher-dimensional physics simulations.
  • Stokes Pair Stability: They identify specific pairs of spaces $(V, Q)$ that are guaranteed to be stable for the Stokes problem, even with high-order global continuity.

Limitations and Future Work

  • Computational Complexity: High-order smooth elements involve a large number of degrees of freedom per tetrahedron. While they offer superior accuracy, the "bookkeeping" and computational cost of solving the resulting linear systems can be significant.
  • Implementation Hurdles: Implementing these elements in standard finite element software is non-trivial due to the complex nature of the degrees of freedom (which include high-order derivatives at vertices and edges).
  • Boundary Conditions: The current work assumes relatively simple domain topologies. Future work could explore how these smooth complexes handle complex boundary conditions or non-simply connected domains (e.g., domains with "holes").

Summary updated

Abstract

Finite element de Rham complexes and finite element Stokes complexes with various smoothness in three dimensions are systematically constructed. First smooth scalar finite elements in three dimensions are derived through a non-overlapping decomposition of the simplicial lattice. Based on the smooth scalar finite elements, both H(div)-conforming finite elements and H(curl)-conforming finite elements with various smoothness are devised, which induce the finite element de Rham complexes with various smoothness and the associated commutative diagrams. The div stability is established for the H(div)-conforming finite elements, and the exactness of these finite element complexes.
BibTeX
@article{chen2023finite,
  author = {Long Chen and Xuehai Huang},
  title = {{Finite element de Rham and Stokes complexes in three dimensions}},
  journal = {Math. Comput.},
  year = {2023},
  doi = {10.1090/mcom/3859},
}

Loading related content...

Figure