Asymptotic gauge-invariant Hybrid High-Order method for magnetic Schrödinger equations

This paper introduces an asymptotic gauge-invariant Hybrid High-Order method for magnetic Schrödinger equations that utilizes a discrete covariant gradient on arbitrary polyhedral meshes to ensure gauge covariance, stability, and optimal convergence, as validated by numerical experiments reproducing key quantum phenomena.

The Gist

The Big Picture: Navigating a Magnetic Maze

Imagine you are trying to simulate a tiny particle (like an electron) moving through a computer model. This particle is moving through a magnetic field. In the real world, physics has a very strict rule: it doesn't matter how you describe the magnetic field, as long as the description is consistent.

Think of the magnetic field like a map. You can draw the map with a grid of squares, or a grid of triangles, or even a grid of hexagons. You can also shift the "zero point" of your map (like changing where you say "North" is). In the real world, the particle doesn't care about your map choices; it only cares about the actual terrain. If you change your map's description, the particle's behavior should stay exactly the same.

The Problem:
When scientists try to simulate this on a computer, they break the world into tiny chunks (like pixels on a screen) to do the math. Most standard computer methods are like a clumsy cartographer: if you change the map description (a "gauge transformation"), the computer simulation gets confused. It starts seeing "ghosts" or fake forces that don't exist in reality. This leads to wrong answers, like predicting the particle has energy it shouldn't have, or behaving erratically over time.

The Solution:
The author, Joubine Aghili, has invented a new, smarter way to do this math called the Hybrid High-Order (HHO) method. It's like giving the computer a "magic compass" that knows how to adjust itself automatically whenever you change the map.

The Key Ingredients

1. The "Magic Compass" (Discrete Covariant Gradient)

In physics, to know how a particle moves in a magnetic field, you need to calculate its "covariant gradient." That's a fancy way of saying: "How is the particle moving, taking into account the magnetic push?"

The author built a special mathematical tool (a discrete covariant gradient) that works on any shape of grid (squares, triangles, weird polyhedrons).

• The Analogy: Imagine you are walking through a forest with a strong wind (the magnetic field). If you just measure your steps, you get the wrong direction because the wind pushes you. This new tool measures your steps and the wind simultaneously. Even if you change how you describe the wind (the gauge), the tool recalculates your path so you still end up in the right place.

2. The "Shape-Shifting" Grid

Most computer models require the grid to be made of perfect squares or triangles. This new method is flexible. It can handle polyhedral meshes (3D shapes with many sides, like a d20 die or a Voronoi cell).

• The Analogy: Imagine building a house. Old methods required you to use only perfect bricks. This new method lets you use rocks, logs, and weirdly shaped stones, as long as they fit together. This is great for complex shapes, like simulating a particle inside a twisted molecule.

3. The "Safety Net" (Gårding Inequality)

In these magnetic simulations, the math can sometimes become unstable, like a ball rolling off a cliff. The author proved that their method has a built-in "safety net" (a discrete Gårding inequality).

• The Analogy: Even if the magnetic field is tricky, this method guarantees that the particle's energy stays within a safe, predictable range. It ensures the simulation doesn't explode or produce nonsense results.

The Proof: Did it Work?

The author didn't just write the theory; they tested it with two famous physics experiments:

Test 1: The Fock-Darwin Spectrum (The Energy Check)
They simulated a particle in a magnetic field and calculated its energy levels.

• The Result: They ran the simulation using three different ways to describe the magnetic field (three different "maps"). In a bad computer model, the energy results would be different for each map. In this new method, the results were identical (within tiny rounding errors). This proves the "magic compass" works perfectly.

Test 2: The Aharonov-Bohm Effect (The Phase Shift)
This is a weird quantum phenomenon where a particle is affected by a magnetic field even if it never touches the field itself (like a ghost passing through a wall).

• The Setup: A particle wave splits into two paths around a magnetic tube and recombines. If the magnetic field is "on," the two waves interfere and create a pattern. If "off," they don't.
• The Result: The simulation perfectly recreated the interference pattern. When the magnetic flux was turned on, the central peak of the wave disappeared (destructive interference), exactly as quantum theory predicts. This shows the method captures the subtle "phase" of the particle correctly.

Why Does This Matter?

This paper is a big deal for two reasons:

1. Accuracy: It stops computers from making "ghost" errors in quantum simulations.
2. Flexibility: It allows scientists to simulate complex 3D shapes (like real-world materials or biological molecules) without being forced to use simple, rigid grids.

In a nutshell: The author built a new, super-flexible, and "gauge-proof" calculator for quantum particles. It ensures that no matter how you describe the magnetic world, the computer tells you the truth about how the particles behave.

Technical Summary

1. Problem Statement

The paper addresses the numerical simulation of the Magnetic Schrödinger Equation, which governs the behavior of non-relativistic quantum particles in the presence of electromagnetic fields. The equation involves a covariant derivative operator $( -i\nabla - \mathbf{A} )$, where $\mathbf{A}$ is the magnetic vector potential.

Key Challenges:

• Gauge Invariance: Physical observables in quantum mechanics are invariant under continuous gauge transformations ($\mathbf{A} \to \mathbf{A} + \nabla \chi$, $\psi \to \psi e^{i\chi}$). Standard discretization schemes often break this symmetry at the discrete level, leading to unphysical artifacts such as spurious eigenvalues and incorrect long-time dynamics.
• Complex Meshes: Existing gauge-invariant methods (e.g., lattice gauge theory or specific Finite Element Methods) are often restricted to uniform grids or simplicial meshes. Extending exact discrete gauge invariance to arbitrary polyhedral meshes with arbitrarily high-order accuracy has remained an open problem.
• Stability: The presence of the magnetic potential destroys the strict coercivity of the standard Schrödinger operator, making the proof of stability (existence of a stable ground state) non-trivial.

2. Methodology: Hybrid High-Order (HHO) Framework

The author proposes a Hybrid High-Order (HHO) method, a modern discretization framework known for its flexibility with arbitrary polyhedral meshes and dimension-independent construction.

Core Components:

1. Discrete Degrees of Freedom (DOFs): The method uses fully discontinuous polynomial spaces on mesh elements ($T$) and faces ($F$). The local DOF space is $U^k_T = P_k(T) \times \prod_{F \subset \partial T} P_k(F)$.
2. Potential Reconstruction: A local reconstruction operator $p^{k+1}_T$ maps the DOFs to a polynomial of degree $k+1$ on the element, satisfying a local Neumann problem.
3. Discrete Covariant Gradient ($G^k_{T, \mathbf{A}}$):
   • This is the central innovation. The author defines a discrete gradient operator that couples the spatial derivative with the magnetic potential.
   • It is constructed via a weak formulation involving the adjoint covariant operator $G^*_\mathbf{A}$ and face jumps.
   • Definition: For a test function $\tau \in P_k(T)^d$,
     $$(G^k_{T, \mathbf{A}} u_{TF}, \tau)_T = (u_T, G^*_\mathbf{A} \tau)_T - i \sum_{F \in \mathcal{F}_T} (u_F, \tau \cdot n_{TF})_F$$
   • This definition ensures the operator respects the coupling between the gradient and the vector potential locally.
4. Stabilization: A standard HHO stabilization term $s_T$ is added to control the jumps between element and face unknowns, ensuring the discrete system is well-posed.

3. Key Contributions and Theoretical Results

A. Asymptotic Discrete Gauge Covariance

The paper proves that the proposed scheme is asymptotically gauge invariant.

• Theorem 3.3: Under a discrete gauge transformation, the discrete covariant gradient transforms covariantly up to a remainder term $R_{gauge}$.
• Error Bound: The remainder is bounded by $O(h^{k+1})$.
• Significance: While exact algebraic gauge invariance is lost due to polynomial projection errors (aliasing), the scheme recovers exact invariance asymptotically as the mesh is refined ($h \to 0$). For low-energy physical states (smooth wavefunctions), the scheme exhibits super-convergence, effectively bypassing the local aliasing limit.

B. Discrete Gårding Inequality

Due to the magnetic potential, the operator is not strictly coercive.

• Theorem 3.8: The author proves a Discrete Gauge-Invariant Gårding Inequality.
• Result: There exist constants $\alpha, \gamma > 0$ such that:
  $$a_h(v_h, v_h) \geq \alpha \|v_h\|_{1,h}^2 - \gamma (\|\mathbf{A}\|_{L^\infty}^2 + \|V\|_{L^\infty}^2) \|v_h\|_{L^2}^2$$
• Implication: This guarantees that the discrete system possesses a stable ground state and a spectrum bounded from below, mirroring the continuous physical properties.

C. Optimal Convergence Rates

• Theorem 3.9: The paper establishes optimal a priori error estimates for the stationary problem.
• Rates:
  • Energy norm: $O(h^{k+1})$
  • $L^2$ norm: $O(h^{k+2})$
• These rates hold for arbitrary polyhedral meshes and high polynomial degrees $k$.

4. Numerical Results

The theoretical findings are validated through two distinct 2D and 3D numerical experiments using the HArDCore3D library.

Experiment 1: Fock-Darwin Spectrum (Eigenvalue Problem)

• Setup: A 2D quantum harmonic oscillator in a uniform magnetic field.
• Goal: Verify gauge invariance and convergence.
• Gauges Tested: Symmetric, Landau, and a manufactured smooth gauge.
• Findings:
  • The computed eigenvalues are independent of the chosen gauge (up to discretization error).
  • The deviation between eigenvalues computed with different gauges scales as $O(h^{k+1})$, confirming the theoretical gauge covariance.
  • Super-convergence: On cubic meshes, the fundamental energy converges at a rate of $O(h^{2k+2})$, significantly better than the standard $O(h^{k+1})$ prediction, attributed to the smoothness of the physical eigenstates.

Experiment 2: Aharonov-Bohm Effect (Time-Dependent Problem)

• Setup: A 3D simulation of a charged particle wavepacket propagating around a solenoid (where $\mathbf{B}=0$ but $\mathbf{A} \neq 0$).
• Goal: Replicate the topological phase shift causing interference patterns.
• Scenarios:
  • Zero Flux ($\Phi=0$): Constructive interference at the center of the detection screen.
  • Flux $\Phi=\pi$: Destructive interference, resulting in a zero-probability node at the center.
• Findings: The HHO scheme successfully reproduces the theoretical interference patterns and the phase shift $\Delta \phi = \Phi$, demonstrating that the discrete covariant gradient correctly captures the topological effects of the magnetic vector potential.

5. Significance and Conclusion

This work represents a significant advancement in computational quantum mechanics by:

1. Bridging the Gap: It extends gauge-invariant discretization to arbitrary polyhedral meshes and high-order methods, a capability previously lacking in the literature.
2. Physical Fidelity: By preserving gauge covariance asymptotically, the method avoids unphysical artifacts, ensuring that numerical solutions respect fundamental quantum mechanical symmetries.
3. Robustness: The proof of the discrete Gårding inequality ensures the method is stable even in the presence of strong magnetic fields where standard coercivity fails.
4. Versatility: The framework is easily extensible to other quantum models (e.g., Pauli equation, Dirac equation) and complex geometries, as demonstrated by the successful 3D Aharonov-Bohm simulation.

The paper concludes that the proposed HHO method is a robust, high-order, and physically consistent tool for simulating quantum systems in magnetic fields, suitable for complex unstructured grids used in modern engineering and physics applications.