Applications of Fourier analysis in homogenization of Dirichlet problem. III: Polygonal domains (Q485247)

From MaRDI portal
scientific article; zbMATH DE number 6409888
  • Applications of Fourier analysis in homogenization of the Dirichlet problem: \(L^p\) estimates
Language Label Description Also known as
English
Applications of Fourier analysis in homogenization of Dirichlet problem. III: Polygonal domains
scientific article; zbMATH DE number 6409888
  • Applications of Fourier analysis in homogenization of the Dirichlet problem: \(L^p\) estimates

Statements

Applications of Fourier analysis in homogenization of Dirichlet problem. III: Polygonal domains (English)
0 references
Applications of Fourier analysis in homogenization of the Dirichlet problem: \(L^p\) estimates (English)
0 references
0 references
0 references
0 references
9 January 2015
0 references
2 March 2015
0 references
homogenization
0 references
Dirichlet problem
0 references
polygonal domain
0 references
Fourier analysis
0 references
elliptic equations in divergence form
0 references
Diophantine condition
0 references
rapidly oscillating boundary data
0 references
convergence rate
0 references
In the paper under review, the authors apply a methods based on Fourier analysis to the investigation of the homogenization of a Dirichlet problem. So let \(D\) be a smooth uniformly convex domain in \(\mathbb{R}^d\) (\(d \geq 2\)), \(A\) a \(\mathbb{R}^{N^2 \times d^2}\)-valued function defined in \(\mathbb{R}^d\) (\(N \geq 1\)), and \(g\) a \(\mathbb{R}^N\)-valued function defined on \(\partial D \times \mathbb{R}^d\). For \(\epsilon>0\) they consider the elliptic system \[ -\mathrm{div}(A(x)\nabla u(x))=0 \qquad \forall x \in D\, , \] with Dirichlet boundary condition \[ u(x)=g(x,x/\epsilon) \qquad \forall x \in \partial D\, . \] They are interested in studying the asymptotic behavior of the solution \(u_\epsilon\) of the boundary value problem above as \(\epsilon \to 0\). So, they denote by \(u_0\) the solution of the same system with Dirichlet boundary condition \[ u(x)=\int_{\mathbb{T}^d}g(x,y)\, dy \qquad \forall x \in \partial D\, , \] where \(\mathbb{T}^d\) is the unit torus in \(\mathbb{R}^d\). They prove that if \(1 \leq p< \infty\) one has \[ \|u_\epsilon-u_0\|_{L^p(D)}\leq C_p \epsilon^{1/2p} \qquad \mathrm{if }\;d=2\, , \] \[ \|u_\epsilon-u_0\|_{L^p(D)}\leq C_p (\epsilon |\ln \epsilon|)^{1/p} \qquad \mathrm{if } \;d=3\, , \] \[ \|u_\epsilon-u_0\|_{L^p(D)}\leq C_p \epsilon^{1/2p} \qquad \mathrm{if } \;d\geq 4\, , \] and they show that such a convergence rate is optimal if \(d\geq 4\). Then they impose periodicity condition on \(A\) and turn to consider the solution \(u_\epsilon\) of the elliptic system \[ -\mathrm{div}(A(x/\epsilon)\nabla u(x))=0 \qquad \forall x \in D\, , \] with the Dirichlet condition \[ u(x)=g(x,x/\epsilon) \qquad \forall x \in \partial D\, . \] Then they denote by \(u_0\) the solution of the homogenized problem and they prove the following convergence rate \[ \|u_\epsilon-u_0\|_{L^p(D)}\leq C_p [\epsilon (\ln (1/\epsilon))^2]^{1/p}\, . \]
0 references
0 references
0 references

Identifiers

0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references