🔢 mathematics

The Landau--Lifshitz--Bloch equation with spin diffusion: Global strong solution and finite element approximation

This paper establishes the existence and uniqueness of global strong solutions for the spin-diffusion Landau--Lifshitz--Bloch equation under small initial data and proposes a decoupled linearised finite element scheme that achieves optimal convergence rates despite the system's strong nonlinearity.

Original authors: Agus L. Soenjaya

Published 2026-04-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Agus L. Soenjaya

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to predict how a crowd of people behaves in a giant, crowded room. In the world of physics, these "people" are tiny magnetic spins inside a material, and the "room" is a computer chip or a memory device.

This paper tackles a very complex problem: How do these tiny magnetic spins behave when they are hot and being pushed around by an electric current?

Here is the breakdown of the paper's story, using simple analogies:

1. The Setting: A Hot, Chaotic Dance Floor

Usually, magnetic materials (like the ones in your hard drive) are cool and orderly. The spins all point in the same direction, like soldiers marching in step. This is described by standard physics equations.

But, modern technology (like heat-assisted magnetic recording) gets these materials very hot—so hot that they exceed their "Curie temperature."

The Analogy: Imagine the soldiers getting so hot they start sweating and losing their formation. They stop marching in a straight line and start wobbling, spinning, and wandering around.
The Problem: When it's this hot, the old math equations break down. We need a new set of rules to describe this "wobbling" behavior. This new set of rules is called the SDLLB equation (Spin-Diffusion Landau-Lifshitz-Bloch).

2. The Two Characters: The Magnet and the Spin Current

The paper models the interaction between two main characters:

The Magnetization ( $m$ ): The collective mood of the crowd (the magnetic field).
The Spin Accumulation ( $s$ ): A stream of "spin-polarized" electrons (like a river of people pushing the crowd).

The Interaction:
The river of electrons pushes the crowd, making them spin and move. At the same time, the crowd's movement affects how the river flows. It's a dance where both partners influence each other constantly.

The Twist: Because the room is so hot, the crowd doesn't just spin; they also lose their energy and eventually stop moving (decay to zero). This is different from the cold scenario where they keep spinning forever.

3. The Challenge: The Math is a Nightmare

The authors' first job was to prove that this "dance" actually makes sense mathematically.

The Difficulty: The equations are nonlinear and coupled.
- Nonlinear: A small change in the push doesn't lead to a small change in the result; it can lead to chaos.
- Coupled: You can't calculate the crowd's movement without knowing the river's flow, and you can't calculate the river without knowing the crowd. It's a "chicken and egg" problem.
The Breakthrough: The authors proved that if the starting crowd isn't too wild (small initial data), the dance will always have a unique, predictable outcome. It won't explode into chaos. They showed that the crowd will eventually calm down and stop, which matches real-world physics.

4. The Solution: A Clever Computer Trick

Now that they knew the math worked, they needed to teach a computer to solve it.

The Problem with Computers: Usually, to solve these coupled equations, a computer has to solve a massive, tangled knot of numbers at every single step. It's like trying to untangle a giant ball of yarn while blindfolded. It takes forever and requires supercomputers.
The Paper's Innovation: The authors invented a new Finite Element Scheme (a method for breaking the problem into tiny, manageable pieces).
- The Magic Trick: They found a way to decouple the problem. Instead of solving the tangled knot, they split it into two separate, simple puzzles.
- The Analogy: Instead of trying to untangle the whole ball of yarn at once, they cut the yarn into two separate strands. They solve one strand, then the other, and they don't interfere with each other.
- The Result: The computer can solve these two simple puzzles simultaneously (in parallel). This makes the simulation incredibly fast and efficient, even for complex 3D shapes.

5. The Proof: Does it Work?

The authors didn't just guess; they did the math to prove their computer method is accurate (converges at an "optimal rate").

The Test: They ran simulations on a computer (using a tool called FEniCS) with different shapes (a disk and a square) and different temperatures.
The Result: The computer simulations matched their theoretical predictions perfectly. The "crowd" behaved exactly as the math said it would: wobbling, interacting with the current, and eventually calming down.

Summary

In short, this paper does three things:

Proves that the physics of hot magnetic materials interacting with electric currents is mathematically sound and predictable.
Invents a super-fast, clever computer method to simulate this physics by splitting a hard problem into two easy ones.
Demonstrates through experiments that this method works perfectly, paving the way for better design of future magnetic memory devices and spintronic computers.

It's like taking a chaotic, hot dance floor, proving the dancers won't crash into each other, and then inventing a new way to film the dance that is 100 times faster than before.

1. Problem Statement

The paper addresses the Spin-Diffusion Landau–Lifshitz–Bloch (SDLLB) system, a nonlinearly coupled system of quasilinear vector-valued partial differential equations (PDEs). This model describes the interaction between magnetization ( $m$ ) and spin accumulation ( $s$ ) in ferromagnetic materials at high temperatures (above the Curie temperature, $T_c$ ), particularly in the presence of spin-polarized currents.

The system is defined on a bounded domain $\Omega \subset \mathbb{R}^d$ ( $d \le 3$ ) and consists of:

Magnetization Equation: A modified Landau–Lifshitz–Bloch equation incorporating a torque term from spin accumulation ( $-\gamma' m \times s$ ). Unlike the standard Landau–Lifshitz–Gilbert (LLG) equation, the LLB equation allows the magnitude of magnetization $|m|$ to vary, which is physically necessary above $T_c$ where thermal agitation causes a loss of macroscopic magnetization.
Spin Diffusion Equation: A nonlinear drift-diffusion equation for spin accumulation $s$ , coupled to the magnetization via a spin-transfer torque term and a diffusion term dependent on $m$ .

Key Challenges:

Nonlinearity: The system features strong nonlinear coupling (e.g., terms like $|m|^2 m$ and $m \otimes m \nabla s$ ).
Lack of Pointwise Bounds: Unlike the LLG equation, the SDLLB system does not automatically preserve the pointwise bound $|m|=1$ , complicating the analysis of global existence.
Well-posedness: Proving the existence of global strong solutions in 3D is difficult due to the potential for backward diffusion behavior if coupling parameters are too large.
Numerical Complexity: Standard numerical approaches for such coupled systems often require solving large, fully coupled nonlinear systems at each time step, which is computationally expensive.

2. Methodology

A. Theoretical Analysis (Existence and Uniqueness)

The author establishes the global existence and uniqueness of strong solutions under specific conditions:

Smallness Assumption: The initial data ( $m_0, s_0$ ) or the spin polarization parameter $\beta$ must be sufficiently small. Specifically, the condition $\beta D_* \|m_0\|_{L^\infty}^2 < D^*$ (where $D_*, D^*$ are bounds on the diffusion coefficient) is required to ensure parabolicity and prevent ill-posedness.
A Priori Estimates: The proof relies on deriving uniform energy estimates using the Faedo–Galerkin method (formally applied to smooth solutions). Key steps include:
- Proving exponential decay of the $L^p$ norms of $m$ and $s$ (Lemma 3.1, 3.2), reflecting the physical loss of magnetization above $T_c$ .
- Establishing higher-order regularity estimates ( $H^2, H^3$ ) using Gagliardo–Nirenberg inequalities and elliptic regularity theory.
- Using a continuous dependence argument (energy-based) to prove uniqueness.

B. Numerical Method (Finite Element Approximation)

The paper proposes a fully-discrete, linearized, decoupled finite element scheme:

Spatial Discretization: Conforming Galerkin finite elements ( $V_h$ ) with polynomial degree $r$ .
Temporal Discretization: Linearized implicit Euler method.
Decoupling Strategy: The scheme treats the nonlinear coupling terms explicitly using values from the previous time step ( $n-1$ $n - 1$ ).
- Step 1: Solve for $m_h^n$ using $m_h^{n-1}$ and $s_h^{n-1}$ .
- Step 2: Solve for $s_h^n$ using $m_h^{n-1}$ and $s_h^{n-1}$ .
- Crucial Advantage: This results in two completely decoupled linear systems per time step, which can be solved in parallel. This contrasts with previous methods for similar systems (like SDLLG) that often require sequential or fully coupled solves.

C. Error Analysis

The convergence analysis involves:

Elliptic Projections: Introducing specialized elliptic projection operators ( $\Pi_h, \Lambda_h$ ) tailored to the bilinear forms of the SDLLB system to handle the variable coefficients and nonlinearities.
Inductive Argument: Proving that if the numerical solution at step $n-1$ satisfies a smallness condition (to ensure coercivity), the solution at step $n$ exists and satisfies the same condition.
Optimal Convergence: Under the assumption of sufficient regularity of the exact solution and a time-step constraint ( $k = O(h^{-d/2})$ $k = O (h^{- d /2})$ ), the scheme is proven to converge with optimal rates:
- $L^2$ error: $O(h^{r+1} + k)$ .
- $H^1$ error: $O(h^r + k)$ .

3. Key Contributions

Global Strong Solution in 3D: The paper extends previous work (which was limited to $d \le 2$ ) by proving the existence and uniqueness of global strong solutions for the SDLLB system in 3D, provided the initial data is sufficiently small.
Efficient Numerical Scheme: It introduces a novel decoupled linearized finite element scheme. Unlike existing methods for spin-diffusion problems that solve coupled nonlinear systems, this method requires solving only two decoupled linear systems per time step, significantly reducing computational cost.
Rigorous Error Estimates: The paper provides a complete error analysis with optimal convergence rates in both space and time. This is a significant advancement over previous works on similar coupled systems (like SDLLG) where only convergence along a subsequence (without rates) was typically proven.
Physical Insight: The analysis rigorously confirms the physical intuition that above the Curie temperature, both magnetization and spin accumulation decay to zero due to thermal agitation and enhanced scattering.

4. Results

Theoretical: Theorem 3.10 establishes the well-posedness of the SDLLB system. Theorem 4.12 proves the stability and optimal convergence of the proposed numerical scheme.
Numerical Experiments: The author implemented the scheme using the FEniCS open-source package. Three simulations were conducted:
1. Disk Magnet ( $T > T_c$ ): Verified the decay of magnetization and spin accumulation.
2. Square Magnet ( $T > T_c$ ): Confirmed the expected order of convergence in both $L^2$ and $H^1$ norms.
3. Below Curie Temperature ( $T < T_c$ ): Demonstrated the scheme's applicability to regimes where magnetization magnitude does not decay to zero (consistent with Ginzburg–Landau theory), showing varying torque and precession.
Convergence Rates: The numerical results corroborated the theoretical prediction of optimal convergence rates (e.g., $O(h^2)$ for linear elements in $L^2$ ).

5. Significance

This work is significant for the field of spintronics and micromagnetics for several reasons:

High-Temperature Modeling: It provides a mathematically rigorous framework for simulating magnetic devices operating at high temperatures (e.g., Heat-Assisted Magnetic Recording - HAMR), where standard LLG models fail.
Computational Efficiency: The proposed decoupled scheme makes large-scale 3D simulations of spin-diffusion effects computationally feasible, which is critical for designing next-generation memory devices (MRAM).
Mathematical Rigor: By establishing optimal error estimates for a highly nonlinear, coupled quasilinear system, the paper sets a new standard for the numerical analysis of spin-diffusion problems, moving beyond weak convergence results to strong, rate-verified convergence.

In summary, the paper successfully bridges the gap between the complex physics of high-temperature spintronics and efficient, provably accurate numerical simulation, offering both theoretical guarantees and practical algorithms for solving the SDLLB system.