English

Singular perturbation of manifold-valued maps with anisotropic energy

Analysis of PDEs 2022-11-16 v2

Abstract

We establish small energy H\"{o}lder bounds for minimizers uεu_\varepsilon of Eε(u):=ΩW(u)+1ε2Ωf(u),E_\varepsilon (u):=\int_\Omega W(\nabla u)+ \frac{1}{\varepsilon^2} \int_\Omega f(u), where WW is a positive definite quadratic form and the potential ff constrains uu to be close to a given manifold N\mathcal N. This implies that, up to subsequence, uεu_\varepsilon converges locally uniformly to an N\mathcal N-valued WW-harmonic map, away from its singular set. We treat general energies, covering in particular the 3D Landau-de Gennes model for liquid crystals, with three distinct elastic constants. Similar results are known in the isotropic case W(u)=u2W(\nabla u)=\vert \nabla u\vert^2 and rely on three ingredients: a monotonicity formula for the scale-invariant energy on small balls, a uniform pointwise bound, and a Bochner equation for the energy density. In the level of generality we consider, all of these ingredients are absent. In particular, the lack of monotonicity formula is an important reason why optimal estimates on the singular set of WW-harmonic maps constitute an open problem. Our novel argument relies on showing appropriate decay for the energy on small balls, separately at scales smaller and larger than ε\varepsilon: the former is obtained from the regularity of solutions to elliptic systems while the latter is inherited from the regularity of WW-harmonic maps. This also allows us to handle physically relevant boundary conditions for which, even in the isotropic case, uniform convergence up to the boundary was open.

Keywords

Cite

@article{arxiv.1809.05170,
  title  = {Singular perturbation of manifold-valued maps with anisotropic energy},
  author = {Andres Contreras and Xavier Lamy},
  journal= {arXiv preprint arXiv:1809.05170},
  year   = {2022}
}

Comments

The initial proof of the energy improvement lemma 2.2 contained a gap and has been corrected in this new version