On the semicontinuity and the relaxation for integrals with respect to the Lebesgue measure added to integrals with respect to a Radon measure (Q1101674)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | On the semicontinuity and the relaxation for integrals with respect to the Lebesgue measure added to integrals with respect to a Radon measure |
scientific article |
Statements
On the semicontinuity and the relaxation for integrals with respect to the Lebesgue measure added to integrals with respect to a Radon measure (English)
0 references
1987
0 references
Let \(\Omega \subseteq {\mathbb{R}}^ n\) be an open set, and consider the functional \(F: H^{1,p}(\Omega)\to {\mathbb{R}}(p\geq 1)\) defined as follows: \[ F(u):=\int_{\Omega}f(x,u(x),Du(x))dx+\int_{{\bar \Omega}}g(x,\tilde u(x))d\mu (x), \] Here \(\tilde u\) is a ``trace'' of u (which coincides with u at continuity points) and \(\mu\) is a nonnegative Radon measure on \({\bar \Omega}\). This paper extends the usual theory of semicontinuity and relaxation (in which \(g\equiv 0)\) to variational problems with obstacles or boundary value problems in which \(g\neq 0\). Thus, the main semicontinuity result (Thm. 5.2) offers conditions on f,g, and \(\Omega\) under which the lower semicontinuity of the first term with respect to \(L^ p(\Omega)\) implies the lower semicontinuity of the functional F itself. These conditions require f(x,s,z)\(\geq \psi(z)\) for some nonnegative convex \(\psi\) satisfying \(\psi (tz)=t^ p\psi(z)\), \(t\geq 0\), \(z\in {\mathbb{R}}^ n\), and that \(g: {\bar \Omega}\times {\mathbb{R}}\to [0,+\infty]\) be a proper normal integrand which is suitably related to both \(\mu\) and \(\Omega\) via capacity theory. The relaxation theorem assumes that \(f(x,s,z)=\psi(z)\) and that f is a Caratheodory function satisfying a certain growth restriction: it offers a formula for the largest \(L^ p(\Omega)\)-lower semicontinuous functional \(\tilde F\) obeying \(\tilde F(u)\leq F(u)\) for all u. More precisely, the formula describes the values of \(\tilde F(u)\) for \(u\in H^{1,p}(\Omega)\cap L^{\infty}(\Omega):\) these are obtained by replacing g(x,s) in the definition of the functional F by a function \(\gamma(x,s)\) described in detail in Thm. 6.1.
0 references
Radon measure
0 references
semicontinuity
0 references
relaxation
0 references
obstacles
0 references
capacity
0 references
0 references