0
$\begingroup$

Let $\Omega \subset \mathbb { R } ^ n $be a bounded domain, and suppose that in the space $W^{ 1,q}(\Omega) $ $ (1 \leq q < \infty)$, the sequence $\{ u_j \} $ converges weakly in $W^{ 1,q}(\Omega) $, and $\{u_j \} $converges almost everywhere to a function $u_0$. My question is whether or not that $\{ u_ j \} $ has a subsequence that converges in $L^q (\Omega)$? If so, converge to what? Is $u_ 0 $ still in $W ^ { 1, q} (\Omega)$?

$\endgroup$
3
  • 2
    $\begingroup$ Is this a question on the regularity of $\Omega$? With some regularity the embedding is compact. $\endgroup$ Commented May 6, 2023 at 11:38
  • $\begingroup$ It follows from Rellich-Kondrachov compactness Theorem, only when $\partial\Omega$ is $C^1$, $1\leq q<n$, $1\leq p<q^*$ with $q^*=\frac{qn}{n-q}$, the embedding $W^{1,q}$ to $L^p$ is compact. Are the assumptions sharp? $\endgroup$ Commented May 7, 2023 at 3:22
  • $\begingroup$ If $\Omega $ is $C^1$, the emebedding of $W^{1,q}$ into $L^q$ is compact for every $q$. See Theorem IX.16 in the book of Brezis. $\endgroup$ Commented May 7, 2023 at 7:33

0

You must log in to answer this question.

Browse other questions tagged .