Anisotropic Curvature Flow and Surface Growth

17 concepts9 papers

In standard Euclidean geometry, a soap bubble is perfectly round because surface tension costs the same in every direction. If you let a shape shrink under its own curvature, it rounds out into a sphere and eventually vanishes. This uniform, direction-independent process is known as isotropic mean curvature flow.

But nature is rarely isotropic. In a crystal lattice, atoms pack tightly along certain planes and loosely along others. Consequently, creating a surface facing one direction might require breaking far more chemical bonds than creating a surface facing another. The energy cost—and by extension, the growth velocity—depends fundamentally on the orientation of the surface normal.

This direction-dependent behavior breaks the symmetry of classical geometric models. To understand how crystals grow, how flames propagate, or how surfaces wrinkle in disordered media, we must replace isotropic metrics with anisotropic ones.

Redefining Perimeter and Surface Area

How do we mathematically measure the area of a surface when space itself has a preferred direction?

A powerful way to build intuition for this is through the concept of Minkowski content. To find the surface area of a shape classically, you imagine wrapping it in a thin "skin" of thickness $\epsilon$ by taking a spherical neighborhood around the boundary. You then measure the volume of this skin and divide it by $\epsilon$ as $\epsilon \to 0$.

In A remark on the anisotropic outer Minkowski content, Stefano Lisini et al. demonstrate how to generalize this concept to anisotropic settings. Instead of inflating the surface using a standard Euclidean ball, the authors inflate it using an arbitrary convex body. The support function of this chosen convex body acts as a directional weight on the surface normal. As the skin shrinks, the resulting measure converges precisely to an anisotropic perimeter. By changing the shape of the inflating neighborhood, we fundamentally alter how different spatial orientations are penalized.

When we look for shapes that minimize this anisotropic perimeter for a given volume, we no longer get spheres; we get the equilibrium crystal shape, known as the Wulff shape. As explored by Mark Green in Steiner polynomials, Wulff flows, and some new isoperimetric inequalities for convex plane curves, modeling physical growth processes requires defining expansion speeds that vary based on the orientation of the surface normal. This shifts the geometric analysis from isotropic Euclidean space to non-Euclidean, norm-dependent contexts.

Evolving Surfaces via Variational Frameworks

Moving from static shapes to dynamic evolution requires defining how a surface moves under its own anisotropic mean curvature. In these flows, the velocity of a point on the surface is determined by how quickly the total anisotropic energy decreases.

Computing this motion directly is notoriously difficult. Anisotropic energies can be non-smooth or "crystalline," causing surfaces to develop sharp facets and corners where standard derivatives fail. To overcome this, researchers turn to variational frameworks that treat the evolution as a sequence of energy minimization problems.

Antonin Chambolle introduced a robust approach to this problem in An algorithm for Mean Curvature Motion by linking curvature motion to Total Variation (TV) minimization. By iteratively solving a minimization problem involving a signed distance function, the algorithm naturally handles topological changes like splitting and merging. This framework extends seamlessly to orientation-dependent motion by substituting Euclidean distance metrics with anisotropic counterparts, allowing simulations to evolve shapes toward their crystalline equilibrium even when the anisotropy is non-smooth.

Real-world materials are rarely perfectly uniform; they are often filled with impurities or spatial variations. In Minimizing movements for anisotropic and inhomogeneous mean curvature flows, Antonin Chambolle et al. generalized this evolution to environments where the cost of surface area depends simultaneously on where the surface is (inhomogeneity) and which way it is facing (anisotropy). Similarly, in Implicit time discretization of the mean curvature flow with a discontinuous forcing term, the authors utilized a Finsler-based cost function to model how direction-dependent surface energies interact with rough, discontinuous external forces.

Furthermore, the geometric stability of these evolving boundaries can be explicitly characterized. In A characterization of convex calibrable sets in ( R^{N} ) with respect to anisotropic norms, Antonin Chambolle et al. proved that certain "calibrable" sets maintain their geometric profiles and evolve uniformly under anisotropic total variation flow, providing a foundational link between geometric stability and anisotropic mean curvature.

Surface Growth in Disordered and Periodic Media

While geometric analysis treats surfaces as smooth or rectifiable sets in continuous space, statistical physics approaches the problem by looking at how surfaces "wrinkle" and roughen at the microscopic scale when driving forces compete with underlying media.

When an interface moves through a periodic medium like a crystal lattice, the energy cost of a minimal surface fluctuates depending on its alignment with the lattice. In Plane-Like Minimizers and Differentiability of the Stable Norm, Antonin Chambolle et al. studied the "stable norm"—the homogenized, macroscopic energy cost of a surface moving through such a medium. They proved that this cost function is smooth in completely irrational directions, but can develop non-differentiable points in rational directions depending on whether the optimal surfaces perfectly fill the space or leave gaps. This illustrates a deep connection between microscopic lattice structure and macroscopic anisotropy.

If the medium is not periodic but disordered (containing random, permanent impurities), the anisotropy manifests as distinct scaling behaviors along different axes. In Renormalization group analysis of the anisotropic Kardar-Parisi-Zhang equation with spatially correlated noise, Hawoong Jeong et al. analyzed the Anisotropic Kardar-Parisi-Zhang (AKPZ) equation. They demonstrated that when growth rates differ along various spatial axes, standard isotropic models fail to capture how these directional differences influence the overall roughening of the interface.

This directional divergence is further illuminated in Anisotropic Surface Growth Model in Disordered Media by Hawoong Jeong et al.. In systems with quenched disorder, surface fluctuations scale differently depending on the orientation relative to the primary growth axis. By calculating separate roughness exponents for the parallel and perpendicular directions, the authors identified a unique universality class where scaling behavior is significantly more pronounced in the direction perpendicular to the growth bias than in the parallel direction.

Whether viewed through the lens of partial differential equations or statistical mechanics, the core insight remains the same: when the energy of a boundary depends on its direction, the resulting evolution fundamentally alters the shape, stability, and roughness of the moving interface.

Key Foundational Work

Anisotropic Surface Growth Model in Disordered Media (1996) — contributed 2 concept elements to this direction.
Renormalization group analysis of the anisotropic Kardar-Parisi-Zhang equation with spatially correlated noise (1995) — contributed 1 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_anisotropic_surface_growth_afa3447e,
  title = {anisotropic surface growth},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/anisotropic-surface-growth/rdr_afa3447ebe9b47dfa190aa780dc9c518},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}

This exact version (version 1, frozen text):

@misc{lacuna_anisotropic_surface_growth_afa3447e_v1,
  title = {anisotropic surface growth},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/anisotropic-surface-growth/rdr_afa3447ebe9b47dfa190aa780dc9c518/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}

See something wrong? Report a mistake