Arbitrary Lagrangian--Eulerian finite element method for lipid membranes

An arbitrary Lagrangian–Eulerian finite element method and numerical implementation for curved and deforming lipid membranes is presented here. The membrane surface is endowed with a mesh whose in-plane motion need not depend on the in-plane flow of lipids. Instead, in-plane mesh dynamics can be specified arbitrarily. A new class of mesh motions is introduced, where the mesh velocity satisfies the dynamical equations of a user-specified two-dimensional material. A Lagrange multiplier constrains the out-of-plane membrane and mesh velocities to be equal, such that the mesh and material always overlap. An associated numerical inf–sup instability ensues, and is removed by adapting established techniques in the finite element analysis of fluids. In our implementation, the aforementioned Lagrange multiplier is projected onto a discontinuous space of piecewise linear functions. The new mesh motion is compared to established Lagrangian and Eulerian formulations by investigating a preeminent numerical benchmark of biological significance: the pulling of a membrane tether from a flat patch, and its subsequent lateral translation.

💡 Research Summary

The paper introduces a fully implicit arbitrary Lagrangian–Eulerian (ALE) finite‑element framework for simulating the coupled in‑plane fluid and out‑of‑plane elastic dynamics of lipid membranes. Traditional approaches either adopt a purely Lagrangian mesh that moves with the material velocity—capturing the membrane flow accurately but suffering severe mesh distortion during large deformations—or an Eulerian mesh that moves only in the normal direction, preserving mesh quality but decoupling the mesh from the material and thus losing geometric fidelity. The authors resolve this dichotomy by allowing the mesh velocity to be governed by an independent two‑dimensional “mesh material” whose governing equations can be chosen arbitrarily by the user (e.g., a simple diffusion‑type PDE, a visco‑elastic law, etc.).

A key physical constraint is that the normal component of the mesh velocity must equal that of the material velocity, ensuring that the mesh always lies on the same surface as the membrane. This constraint is enforced with a scalar Lagrange multiplier p(ζ,t) that acts as a normal pressure per unit area. Because p couples a scalar field to vector velocities, the resulting mixed formulation violates the classical inf‑sup (LBB) condition, leading to spurious oscillations in the pressure‑like field. To stabilize the system, the authors adopt the Dohrmann–Bochev technique: the multiplier is projected onto a discontinuous piecewise‑linear space (denoted (\tilde L)), and a quadratic penalty term penalizing deviations from this projection is added to the weak form. This approach, originally developed for pressure stabilization in incompressible flow, provides a mathematically sound and easy‑to‑implement remedy.

The governing equations comprise: (1) an area‑incompressibility condition (∇·v = 0) enforced by a surface tension λ, (2) the Helfrich bending energy expressed through mean curvature H and Gaussian curvature K, (3) in‑plane viscous stresses proportional to the 2‑D membrane viscosity ζ, and (4) the user‑specified mesh dynamics. The weak forms are derived on a parametrized surface S(ζ₁,ζ₂,t) with basis vectors aα and unit normal n. Function spaces are chosen as continuous H¹ for positions and velocities, continuous L² for λ, and discontinuous L² for p.

Implementation details are provided for a Julia package, MembraneAleFem.jl. The surface is discretized with triangular elements; velocity and position fields use standard continuous shape functions, while the Lagrange multiplier uses discontinuous linear shape functions. Time integration employs a fully implicit Newton–Raphson scheme, and automatic differentiation in Julia supplies consistent Jacobians. The code is open‑source, facilitating reproducibility and extension.

The authors validate the method with a biologically relevant benchmark: pulling a tubular tether from a flat membrane patch and subsequently translating the tether laterally. Three mesh strategies are compared: pure Lagrangian, pure Eulerian, and the proposed ALE. All three produce similar force‑extension curves during the initial pulling phase, but divergences appear as the tether elongates and the surrounding membrane deforms substantially. The Lagrangian mesh collapses due to extreme element distortion, causing solver failure. The Eulerian mesh, while remaining well‑shaped, fails to capture the evolving curvature accurately, leading to erroneous force predictions. In contrast, the ALE mesh maintains element quality through the user‑defined mesh dynamics and respects the normal‑velocity constraint, yielding stable and accurate force‑extension data and realistic tether translation velocities. Additional tests with different mesh‑material constitutive laws demonstrate that the user can tune mesh behavior (e.g., stiffness, diffusion) to balance computational cost and accuracy.

In summary, the paper makes three principal contributions: (i) a general ALE formulation that decouples mesh motion from material flow while preserving geometric consistency, (ii) a robust stabilization of the mixed velocity‑pressure system via Dohrmann–Bochev projection, and (iii) an open‑source Julia implementation that enables high‑performance simulations of complex membrane mechanics. Limitations include the current focus on single‑component, area‑incompressible membranes and the omission of fully coupled 3‑D surrounding fluid dynamics. Future work is suggested in extending the framework to multi‑component phase separation, coupling with bulk Stokes flow, and incorporating adaptive mesh refinement or automatic remeshing strategies.