Second order smoothness of weak solutions of degenerate linear equations (Q1745643): Difference between revisions
From MaRDI portal
Set OpenAlex properties. |
Created claim: Wikidata QID (P12): Q130097818, #quickstatements; #temporary_batch_1730843543283 |
||
(One intermediate revision by one other user not shown) | |||
Property / cites work | |||
Property / cites work: Q5848416 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q3129869 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Isoperimetric and Sobolev inequalities for Carnot-Carath�odory spaces and the existence of minimal surfaces / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q4821332 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Degenerate Sobolev spaces and regularity of subelliptic equations / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q5244856 / rank | |||
Normal rank | |||
Property / Wikidata QID | |||
Property / Wikidata QID: Q130097818 / rank | |||
Normal rank |
Latest revision as of 22:52, 5 November 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Second order smoothness of weak solutions of degenerate linear equations |
scientific article |
Statements
Second order smoothness of weak solutions of degenerate linear equations (English)
0 references
18 April 2018
0 references
Consider an equation \[ \sum_{k=1}^m Y_k' Y_k u = b \] on a bounded open \(n\)-dimensional Euclidean set. The author finds conditions under which the solution \(u\) has weak second order derivatives \(ZY_k u\) and \(Y_kZu\) in \(L^2\). Here \(Y_k\) and \(Z\) are first order vector field derivatives, \(Z = U \cdot \nabla\) with \(U \in C^{1,1}\) and \(Y_k\) share the same regularity. More precise assumptions are that \(b \in L^2\) with a weak derivative \(Zb \in L^2\), and the solution \(u\) has to have weak \(Y\)-derivatives in \(L^2\). Furthermore, there needs to exist a sequence of smooth functions \(u_l\), which converges to \(u\) in terms of the \(H^{1,2}_Y\) norm with the weak vector field derivative \(Y\), and which satisfies \[ \limsup_{l \to \infty} \left( \int \left| \left\{ \sum_{k} |W_k(x)| + U(x) \right\}\nabla u_l(x) \right|^2 d x \right)^{1/2} < \infty. \] Here \(U\) is defined by \(Z = U \cdot \nabla\) and \(W\) by \(Y = W \cdot \nabla\). There is also a local control of the \(L^2\) norm of \(ZY_k u\) in terms of the above limit superior. As a corollary one has a similar result for \(Y_kZu\). Also, for the solutions \(u\) of the linear equations \[ \nabla \cdot (Q\nabla u) = b, \] where the rows of the square root of the symmetric non-negative definite matrix \(Q\) are Lipschitz vectors, one gets that the iterated derivatives \(ZX_iu\) and \(X_i Z u\) exist under similar assumptions as before, and when \(X\) is related to the square root of the matrix \(Q\). As Remark 2.5 the paper contains an erratum to the autor's paper [Acta Math. Sin., Engl. Ser. 34, No. 1, 42--62 (2018; Zbl 1393.35054)], stating that the global result given therein is actually local.
0 references
degenerate elliptic partial differential equations
0 references
degenerate quadratic forms
0 references
weak solutions
0 references
second order regularity
0 references
0 references