Arbitrary Lagrangian-Eulerian Mapping

• ALE mapping is a flexible computational framework that decouples mesh motion from material flow, ensuring stability in simulations on moving domains.
• It transforms partial differential equations into formulations on deforming meshes while rigorously enforcing the geometric conservation law.
• The method underpins modern finite element, finite volume, and high-order schemes, crucial for fluid-structure interaction and evolving boundary problems.

Arbitrary Lagrangian-Eulerian (ALE) Mapping

The Arbitrary Lagrangian-Eulerian (ALE) mapping is a fundamental geometric and algorithmic construct in computational mechanics, enabling the numerical solution of partial differential equations on time-dependent domains that undergo complex, possibly large, motion and deformation. ALE techniques generalize both purely Lagrangian (material point-following) and Eulerian (fixed spatial) descriptions by allowing the computational mesh to move independently of the material, providing flexibility to optimize mesh quality and stability during simulations. This framework is central to modern finite element, finite volume, meshless, and interface-tracking approaches for moving boundary, fluid-structure interaction, free-surface, and evolving surface problems, and is essential for the satisfaction of the geometric conservation law (GCL), which guarantees physically consistent and numerically stable mesh motion.

1. Mathematical Foundation of ALE Mappings

The ALE approach introduces a time-dependent mapping between a reference "material" or computational domain and the current, possibly moving, physical domain:

• For domains in $\mathbb{R}^d$, define a map $\phi(\hat x, t): \Omega_0 \to \Omega_t$, where $\hat x$ are reference coordinates and $\Omega_t$ is the deformed mesh at time $t$ (Ivancic et al., 2018, Shen et al., 2 Feb 2026).
• The mesh (grid) velocity is given by $$w(x, t) = \partial_t \phi(\hat x, t)\rvert_{\hat x = \phi^{-1}(x, t)}$$, distinguishing the motion of the computational grid from the material velocity in the governing physics (Sauer, 29 Sep 2025).

ALE schemes typically rely on the explicit construction of the Jacobian and metric tensors associated with $\phi$:

• The Jacobian $J = \det(\nabla_{\hat x} \phi)$ maps volume elements, is central to volume-conserving updates, and its evolution encodes the discrete or continuous form of GCL.
• On evolving manifolds, the ALE map is constructed in curvilinear coordinates $\zeta^{\alpha}$, generating the instantaneous parametrization $x = \chi(\zeta, t)$ with metric tensor $a_{\alpha\beta}$ and surface Jacobian $J_a = \sqrt{\det [a_{\alpha\beta}]}$ (Sauer, 27 Feb 2025, Sauer, 29 Sep 2025).

The mapping may be constructed using isoparametric basis functions—interpolating both geometry and solution variables—and extended in spacetime for high-order accuracy (e.g., in ADER or DG schemes) (Boscheri et al., 2014, Boscheri et al., 2013).

2. ALE Mapping in Governing Equations and Discretization

The ALE framework systematically transforms PDEs with moving domains into forms suitable for computation on deforming meshes. The differential form for a conserved field $u$ under ALE mapping is: $$\frac{\partial u}{\partial t}\Big|_{\text{mesh}} + \nabla \cdot [(v - w) u] = S(u)$$ where $w$ is the mesh velocity and $v$ the material velocity (Schwarzmeier et al., 2024). The spatial and temporal derivatives in the physical domain are expressed via the ALE map, reparametrizing integrals and derivatives to the moving mesh:

• Volume integrals are transformed using the Jacobian: $$\int_{\Omega_t} f(x, t) dx = \int_{\Omega_0} f(\phi(\hat x, t), t) J(\hat x, t) d\hat x$$.
• Time derivatives at fixed material points are related to mesh derivatives via $$\partial_t f|_{\hat x} = \partial_t f|_{x} + w \cdot \nabla f$$ (Ivancic et al., 2018).
• In surface PDEs, the ALE velocity decomposes into tangential and normal contributions, with the normal part typically constrained by kinematic compatibility, while in-plane motion is arbitrary or subject to elasticity, Laplacian or smoothing PDEs (Sauer, 27 Feb 2025, Sauer, 29 Sep 2025).

For finite element (FEM), finite volume (FVM), and DG-ALE schemes, this mapping is manifested either by elemental isoparametric transformations, reference-to-physical coordinate changes, or in the assembly of differential operators with mesh-velocity corrections.

3. Geometric Conservation Law and Discrete Compatibility

The geometric conservation law (GCL) is intrinsic to ALE frameworks and ensures that mesh-induced volume changes (or area changes on surfaces) do not introduce artificial sources or sinks in the numerical scheme:

• The continuous form is $$d/dt \int_{E(t)} dV - \int_{\partial E(t)} w \cdot n\, dS = 0$$ for a control volume $E(t)$ (Sahin et al., 2010).
• Discrete GCL is enforced by exact algebraic identities relating mesh volume (or area) change to integrated mesh-velocity flux across faces (Gaburro et al., 2024, Zhang et al., 2024). For example, in finite volume schemes: $$|E^{n+1}| - |E^n| = \Delta t \sum_{f} (w \cdot n)_f |S_f|$$ which must be satisfied at each cell update to avoid mesh-pumping errors (Sahin et al., 2010, Colombo et al., 2022).
• For high-order time integration, temporal polynomials for the ALE map enable exact quadrature of mesh-velocity terms, so discrete SCL is satisfied to machine precision for polynomials of any degree in time (Ivancic et al., 2018).
• On evolving manifolds, surface area conservation is ensured by stretching factors and the metric tensor evolution (Sauer, 27 Feb 2025).

Consistent enforcement of GCL is essential for long-term stability, especially in mesh-adaptive or topologically-evolving codes, and is a defining property of robust ALE schemes.

4. Mesh Motion Strategies and ALE Velocity Construction

A central feature of ALE mapping is the freedom to prescribe or solve for the mesh velocity $w$ independently from the material velocity. Practically, three principal strategies are prevalent:

• Elasticity/Laplace Smoothing: The mesh velocity is determined as the solution of a PDE, e.g., $\Delta w = 0$ with Dirichlet boundary data on moving interfaces, ensuring smooth interior mesh motion and preserving quality (Ammad et al., 16 Jan 2026, Shen et al., 2 Feb 2026).
• Optimization: Mesh nodes are moved to minimize mesh distortion or optimize shape metrics, sometimes via mechanical analogies like spring networks on surfaces (Kovács, 2016), or energy minimization (Sauer, 27 Feb 2025).
• Direct Lagrangian or Hybrid Motion: Nodes follow the local fluid velocity (pure Lagrangian) or a combination of Lagrangian and prescribed components, possibly blended with smoothing or rezoning to inhibit tangling (Gaburro et al., 2024, Boscheri et al., 2014).

Boundary motion may be explicitly prescribed (interface-tracking), obtained from boundary physics (e.g., fluid-structure coupling), or driven by curvature, as in surface tension flows. Various meshless or mesh adaptation techniques (e.g., MFS mesh-motion, B-spline geometric reconstruction) have been demonstrated to generate high-quality moving meshes especially in large-deformation or nonconvex domains (Ammad et al., 16 Jan 2026).

5. ALE Mapping in High-Order and Unfitted Methods

The flexibility of ALE mapping underpins numerous modern high-order Eulerian–Lagrangian numerical frameworks:

• Isoparametric Space-Time Mappings: Used in ADER and high-order DG/FV schemes, the same basis polynomials define both solution and geometry, enabling arbitrary high-order in space and time, even on moving, possibly nonconforming, unstructured meshes (Boscheri et al., 2014, Gaburro et al., 2024, Boscheri et al., 2013).
• Unfitted/Embedded FE Methods: Unfitted ALE frameworks map from reference to physical domains, supporting moving boundaries and interfaces without remeshing, and enabling topological changes (Lu et al., 2024).
• Surface ALE: ALE strategies have been extended to evolving surfaces and manifolds, with the mesh velocity constructed to satisfy kinematic, in-plane elasticity, or smoothing constraints, supporting simulations of fluidic interfaces, membranes, and complex surface dynamics (Sauer, 29 Sep 2025, Sauer, 27 Feb 2025).
• Interface-tracking in Multiphase CFD: In codes such as twoPhaseInterTrackFoam, ALE mapping, in conjunction with interface-tracking and subgrid-scale closures, allows for robust, mass-conserving integration of multiphase Navier-Stokes with moving free boundaries (Schwarzmeier et al., 2024).

These high-order and unfitted ALE implementations often feature built-in mechanisms for automatic satisfaction of GCL and are validated for order-optimal convergence, preservation of equilibria, and geometric robustness even under strong deformation or topology change (Lu et al., 2024, Sauer, 29 Sep 2025).

6. Implementation Examples and Algorithmic Patterns

ALE mapping frameworks are now foundational in diverse application domains, including but not limited to:

• Moving boundary flows in cylindrical or axisymmetric geometries, where the governing equations are adapted (e.g., multiplied with radial distance in swirl-free coordinates) to ensure simplicity and discrete exactness (Sahin et al., 2010).
• Fluid–structure interaction with localized mesh motion and partitioned/monolithic time stepping, as in monolithic localized high-order ALE-FEM with IMEX-PRK schemes (Shen et al., 2 Feb 2026).
• Nonconforming ALE schemes using space-time conservation forms that handle sliding interfaces and prevent mesh tangling in shear-dominated flows (Gaburro et al., 2016).
• Particle-laden two-phase flows, using ALE-based tracking and conservative remapping to couple Lagrangian particles and continuum fields efficiently, while enforcing GCL (Zhang et al., 2024).

The selection of mesh velocity, the enforcement of GCL, and the design of algorithms for mesh update, connectivity change, and variable remapping are critical to the correctness and efficiency of each of these implementations.

7. Assessment, Limitations, and Performance

ALE mapping is a mature, mathematically rigorous strategy underpinning state-of-the-art moving-mesh numerical PDE solvers. Robustness with respect to mesh quality, long-time conservation, and optimal accuracy is well established when GCL is enforced; high-order time and space accuracy is possible with isoparametric or polynomial-in-time mappings (Ivancic et al., 2018, Gaburro et al., 2024). ALE frameworks readily accommodate large deformations, complex boundary motions, and mesh adaptation, outperforming Lagrangian-only or fixed-mesh Eulerian methods in many challenging benchmark problems (Sauer, 29 Sep 2025, Ammad et al., 16 Jan 2026).

However, challenges remain in automated handling of topological changes, dynamic remeshing, and computational cost of frequent mesh updates in very large simulations. Extensions to high-speed advection, severe boundary distortion, or fully coupled multiphysics domains (e.g., interface tracking in two-phase ejecta modeling (Zhang et al., 2024)) continue to drive methodological developments.

In summary, ALE mapping provides a unifying and flexible methodology for mesh and domain evolution in computational mechanics and numerical PDEs, with rigorous treatment of geometric conservation, mesh motion, and high-order temporal and spatial discretization (Sahin et al., 2010, Gaburro et al., 2024, Sauer, 29 Sep 2025, Shen et al., 2 Feb 2026).