<p>We study the incompressible magnetohydrodynamic system endowed with a Caputo time-fractional derivative of order <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(0&lt;\alpha &lt;1\)</EquationSource> </InlineEquation>. First, we reformulate the coupled momentum-induction equations by applying a double-curl projection that eliminates both the hydrodynamic pressure and the magnetic pseudo-pressure, producing a divergence-free velocity-magnetic field pair. Next, we then discretise the reformulated problem by combining a divergence-free Fourier spectral approximation in space with a variable-step L1 convolution discretisation in time. Our fully discrete method satisfies a discrete fractional kinetic–magnetic energy inequality, enforces the divergence constraints to machine precision, and reduces to the classical magnetohydrodynamic energy law as <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(\alpha \rightarrow 1^{-}\)</EquationSource> </InlineEquation>, thereby ensuring asymptotic compatibility. Using discrete orthogonal-complementary convolution identities together with discrete Sobolev embeddings, we derive optimal error bounds of order <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(O\!\bigl (\tau ^{2-\alpha }+N^{-m}\bigr )\)</EquationSource> </InlineEquation> on arbitrarily graded time meshes. Finally, we present numerical experiments, including fractional magnetic Taylor–Green and Orszag–Tang vortices, which confirm the theoretical convergence rates, demonstrate monotone energy decay, and highlight the efficiency of the adaptive time-stepping strategy.</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

Energy-dissipative adaptive-step L1 discretisation for the Caputo time-fractional incompressible magnetohydrodynamic system

  • Muhammad Zainul Abidin

摘要

We study the incompressible magnetohydrodynamic system endowed with a Caputo time-fractional derivative of order \(0<\alpha <1\) . First, we reformulate the coupled momentum-induction equations by applying a double-curl projection that eliminates both the hydrodynamic pressure and the magnetic pseudo-pressure, producing a divergence-free velocity-magnetic field pair. Next, we then discretise the reformulated problem by combining a divergence-free Fourier spectral approximation in space with a variable-step L1 convolution discretisation in time. Our fully discrete method satisfies a discrete fractional kinetic–magnetic energy inequality, enforces the divergence constraints to machine precision, and reduces to the classical magnetohydrodynamic energy law as \(\alpha \rightarrow 1^{-}\) , thereby ensuring asymptotic compatibility. Using discrete orthogonal-complementary convolution identities together with discrete Sobolev embeddings, we derive optimal error bounds of order \(O\!\bigl (\tau ^{2-\alpha }+N^{-m}\bigr )\) on arbitrarily graded time meshes. Finally, we present numerical experiments, including fractional magnetic Taylor–Green and Orszag–Tang vortices, which confirm the theoretical convergence rates, demonstrate monotone energy decay, and highlight the efficiency of the adaptive time-stepping strategy.