Gradient weighted estimates at the natural exponent for quasilinear parabolic equations

From MaRDI portal
Publication:2631944



Abstract: In this paper, we obtain weighted norm inequalities for the spatial gradients of weak solutions to quasilinear parabolic equations with weights in the Muckenhoupt class Afracqp(mathbbRn+1) for qgeqp on non-smooth domains. Here the quasilinear nonlinearity is modelled after the standard p-Laplacian operator. Until now, all the weighted estimates for the gradient were obtained only for exponents q>p. The results for exponents q>p used the full complicated machinery of the Calder'on-Zygmund theory developed over the past few decades, but the constants blow up as qightarrowp (essentially because the Maximal function is not bounded on L1). In order to prove the weighted estimates for the gradient at the natural exponent, i.e., q=p, we need to obtain improved a priori estimates below the natural exponent. To this end, we develop the technique of Lipschitz truncation based on cite{AdiByun2,KL} and obtain significantly improved estimates below the natural exponent. Along the way, we also obtain improved, unweighted Calder'on-Zygmund type estimates below the natural exponent which is new even for the linear equations.


The paper deals with Calderón-Zygmund type regularity results in weighted Lebesgue spaces for parabolic equations of the form \[ \begin{cases} u_t-\mathrm{div}\,\mathcal{A}(x,t,\nabla u)= \mathrm{div}\,|\mathbf{f}|^{p-2}\mathbf{f} & \text{in}\ \Omega_T:=\Omega\times (-T,T),\\ u=0 & \text{on}\ \partial_p\Omega_T, \end{cases} \] where the nonlinearity $\mathcal{A}(x,t,\nabla u)$ is modeled after the $p$-Laplacian with $p>\frac{2n}{n+2},$ $\Omega\subset \mathbb{R}^n$ is a bounded domain with possibly non-smooth boundary, and $\partial_p\Omega_T$ is the parabolic boundary of the cylinder $\Omega_T.$ \par The authors prove two type of results. The first one asserts that \[ \mathbf{f}|\in L^q(\Omega_T)\ \Longrightarrow \ |\nabla u|\in L^q(\Omega_T)\quad \forall q\in [p-\beta_0,p] \] with the corresponding estimate, where $\beta_0$ is a small enough universal constant. The importance of this result is twofold: first, because the gradient estimate is obtained below the natural exponent $p$ and second, since no regularity of the coefficients is assumed which makes the the estimate non-perturbative. The technique employed is based on the Lipschitz truncation approach developed in [\textit{J. Kinnunen} and \textit{J. L. Lewis}, Ark. Mat. 40, No. 1, 105--132 (2002; Zbl 1011.35039)]. \par The second type of results establishes the implication \[ \mathbf{f}|\in L^q_\omega(\Omega_T)\ \Longrightarrow \ |\nabla u|\in L^q_\omega(\Omega_T)\quad \forall q\geq p\quad \text{and}\quad \omega\in A_{q/p}(\mathbb{R}^{n+1}) \] between weighted Lebesgue spaces with a Muckenhoupt weight $\omega.$ The novelty of the weighted estimates derived consists in the treatment of the limit case of the natural exponent $q=p.$



Cites work









This page was built for publication: Gradient weighted estimates at the natural exponent for quasilinear parabolic equations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2631944)