Asymptotic behavior of minimizers for the Ginzburg-Landau functional with weight. I (Q1390776)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Asymptotic behavior of minimizers for the Ginzburg-Landau functional with weight. I |
scientific article |
Statements
Asymptotic behavior of minimizers for the Ginzburg-Landau functional with weight. I (English)
0 references
11 January 1999
0 references
Let \(G\) be a bounded, simply connected, smooth domain in \({\mathbf R}^2\), \(g:\partial G\rightarrow S^1\) a smooth boundary data of topological degree \(d>0\) and \(p:\overline{G}\rightarrow (0,+\infty)\) a smooth function. The authors study the asymptotic behavior, as \(\varepsilon\rightarrow 0\), of the minimizers for the Ginzburg-Landau functional \[ E_\varepsilon (u)={2}^{-1}\int_Gp|\nabla u|^2+{4\varepsilon^2}^{-1} \int_G(1-| u|^2)^2 \] on the set \(H^1_g=\{ u\in H^1(G,{\mathbf R}^2\); \(u=g\) on \(\partial G\}\). The functional \(E_\varepsilon\) is related to the Ginzburg-Landau energy in superconductivity and the presence of the nonconstant weight \(p\) is motivated by the problem of pinning of vortices, that is, of forcing the location of the vortices to some restricted sites. Set \(p_0=\min\{p(x)\); \(x\in\overline {G}\}\) and \(\Lambda =p^{-1}(p_0)\). The main result of this paper establishes the convergence of a subsequence \(u_{\varepsilon_n}\rightarrow u_*\) in \(C^{1,\alpha}_{\text{loc}} (\overline{G}\setminus\{b_1,\dots ,b_N\},S^1)\) for every \(\alpha <1\), where the \(N\) distinct points \(\{b_1,\dots ,b_N\}\) lie in \(\Lambda\) and \(u_*\in C^\infty (\overline{G}\setminus\{b_1,\dots ,b_N\},S^1)\) is a solution of \[ -\text{div} (p\nabla u_*)= p|\nabla u_*|^2u_*\quad \text{in} \overline{G}\setminus\{ b_1,\dots ,b_N\},\qquad u_* =g\quad \text{on} \partial G. \] Moreover, the topological degree \(d_k\) of \(u_*\) around each singularity \(b_k\) satisfies \(d_k\geq 1\) and \(\sum_{k=1}^Nd_k=d\) and, for \(n\) sufficiently large, every minimizer \(u_{\varepsilon_n}\) has exactly \(d\) zeros. The paper generalizes in a non trivial framework several results obtained in the last few years by Bethuel, Brezis, Hélein and Struwe for the ``flat'' case \(p\equiv 1\). [See also the following review of Part II].
0 references
minimization problem
0 references
Ginzburg-Landau energy
0 references
singularity
0 references
pinning effect
0 references
winding number
0 references