<p>We consider the Dirichlet problem of the indefinite Helmholtz equation in 1D, <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(u''+k^2u=f\)</EquationSource> </InlineEquation> in (0,&#xa0;1), <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(u(0)=g_0\)</EquationSource> </InlineEquation>, <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(u(1)=g_1\)</EquationSource> </InlineEquation>, with a constant wavenumber <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(k\in (0,\infty )\backslash \pi \mathbb {N}\)</EquationSource> </InlineEquation> and a source term <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(f\in H^p_0(0,1)\)</EquationSource> </InlineEquation>, <InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(p\ge 4\)</EquationSource> </InlineEquation>. We propose an approach based on Fourier analysis to derive wavenumber explicit sharp estimates of absolute and relative errors of <i>finite difference</i> methods. Such results have been well known for <i>finite element</i> methods (FEM). We use the approach to analyze the classical centered finite difference scheme. For the Fourier interpolants of the discrete solution with homogeneous (or inhomogeneous) Dirichlet conditions, we show rigorously, under the two assumptions <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\(k&gt;20\)</EquationSource> </InlineEquation> and <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\(k(kh)^2/\sigma _k\le 4/(\pi -2)\)</EquationSource> </InlineEquation> with <InlineEquation ID="IEq9"> <EquationSource Format="TEX">\(\sigma _k:={\text {dist}}(k,\pi \mathbb {N})\)</EquationSource> </InlineEquation>, that the worst case attainable convergence order of the absolute error with <InlineEquation ID="IEq10"> <EquationSource Format="TEX">\(\sum _{p=0}^4k^{-p}\Vert f^{(p)}\Vert _{L^2}=O(1)\)</EquationSource> </InlineEquation> (or <InlineEquation ID="IEq11"> <EquationSource Format="TEX">\(|g_i|\asymp k^{-1}\)</EquationSource> </InlineEquation>) is <InlineEquation ID="IEq12"> <EquationSource Format="TEX">\({(kh)^2/\sigma _k^2}\)</EquationSource> </InlineEquation> in the <InlineEquation ID="IEq13"> <EquationSource Format="TEX">\(L^2\)</EquationSource> </InlineEquation>-norm and <InlineEquation ID="IEq14"> <EquationSource Format="TEX">\({k(kh)^2/\sigma _k^2}\)</EquationSource> </InlineEquation> in the <InlineEquation ID="IEq15"> <EquationSource Format="TEX">\(H^1\)</EquationSource> </InlineEquation>-semi-norm, and that of the relative error is <InlineEquation ID="IEq16"> <EquationSource Format="TEX">\({k(kh)^2/\sigma _k}\)</EquationSource> </InlineEquation> in both <InlineEquation ID="IEq17"> <EquationSource Format="TEX">\(L^2\)</EquationSource> </InlineEquation>- and <InlineEquation ID="IEq18"> <EquationSource Format="TEX">\(H^1\)</EquationSource> </InlineEquation>-semi-norms if <InlineEquation ID="IEq19"> <EquationSource Format="TEX">\(\Vert u^{(p)}\Vert _{L^2}/\Vert u^{(p-2)}\Vert _{L^2}\asymp k^2\)</EquationSource> </InlineEquation> for <InlineEquation ID="IEq20"> <EquationSource Format="TEX">\(p=2,3\)</EquationSource> </InlineEquation>. In particular, the lower bounds of these error estimates are established rigorously in the same orders as the upper bounds, which is the main novelty of this work. We show also that the Fourier analysis approach can be used as a convenient visual tool for evaluating finite difference schemes in presence of source terms, which is beyond the scope of dispersion analysis. The results from the theory and visual analysis are corroborated by numerical experiments.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Fourier analysis of finite difference schemes for the Helmholtz equation in 1D with Dirichlet conditions: Sharp estimates and relative errors

  • Martin J. Gander,
  • Hui Zhang,
  • Haiyang Zhou

摘要

We consider the Dirichlet problem of the indefinite Helmholtz equation in 1D, \(u''+k^2u=f\) in (0, 1), \(u(0)=g_0\) , \(u(1)=g_1\) , with a constant wavenumber \(k\in (0,\infty )\backslash \pi \mathbb {N}\) and a source term \(f\in H^p_0(0,1)\) , \(p\ge 4\) . We propose an approach based on Fourier analysis to derive wavenumber explicit sharp estimates of absolute and relative errors of finite difference methods. Such results have been well known for finite element methods (FEM). We use the approach to analyze the classical centered finite difference scheme. For the Fourier interpolants of the discrete solution with homogeneous (or inhomogeneous) Dirichlet conditions, we show rigorously, under the two assumptions \(k>20\) and \(k(kh)^2/\sigma _k\le 4/(\pi -2)\) with \(\sigma _k:={\text {dist}}(k,\pi \mathbb {N})\) , that the worst case attainable convergence order of the absolute error with \(\sum _{p=0}^4k^{-p}\Vert f^{(p)}\Vert _{L^2}=O(1)\) (or \(|g_i|\asymp k^{-1}\) ) is \({(kh)^2/\sigma _k^2}\) in the \(L^2\) -norm and \({k(kh)^2/\sigma _k^2}\) in the \(H^1\) -semi-norm, and that of the relative error is \({k(kh)^2/\sigma _k}\) in both \(L^2\) - and \(H^1\) -semi-norms if \(\Vert u^{(p)}\Vert _{L^2}/\Vert u^{(p-2)}\Vert _{L^2}\asymp k^2\) for \(p=2,3\) . In particular, the lower bounds of these error estimates are established rigorously in the same orders as the upper bounds, which is the main novelty of this work. We show also that the Fourier analysis approach can be used as a convenient visual tool for evaluating finite difference schemes in presence of source terms, which is beyond the scope of dispersion analysis. The results from the theory and visual analysis are corroborated by numerical experiments.