Some observations on the regularity of the solutions of strongly nonlinear problems and applications (Q1092322)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Some observations on the regularity of the solutions of strongly nonlinear problems and applications
scientific article

    Statements

    Some observations on the regularity of the solutions of strongly nonlinear problems and applications (English)
    0 references
    0 references
    0 references
    1985
    0 references
    The authors study at first the regularity of solutions of the strongly nonlinear elliptic problem \[ u\in W_ 0^{1,p}(\Omega),\quad g(.,u)\in L^ 1(\Omega),\quad ug(.,u)\in L^ 1(\Omega), \] \[ <Au- f,v>+\int_{\Omega}g(.,u)v=0;\quad \forall v\in W_ 0^{1,p}(\Omega)\cap L^{\infty}(\Omega)\quad and\quad for\quad v=u; \] where \(\Omega\) is a bounded domain in \(R^ N\), A is a strongly nonlinear elliptic operator of second order and growth p \((p>1)\) and g is a Carathéodory function satisfying one-sided condition \(g(x,\eta)\eta \geq 0\) and sup\(\{g(x,\eta)\); \(| \eta | <\lambda\} = h_{\lambda}(x)\in L^ 1(\Omega).\) Let \(f=-\sum (f_ i)_{x_ i}\), \(f_ i\in L^ q(\Omega)\), \(p\leq N\). It is proved that for \(q>N/(p-1)\) the solution u is bounded on \(\Omega\) and for \(p'\leq q<N/(p-1)\) the solution u belongs to \(L^ s(\Omega)\) where s is the exponent of Sobolev embedding theorem corresponding to q(p-1). Under additional conditions on g, the authors prove higher integrability of g(u). Moreover, they extend the previous results to variational inequalities with an obstacle \(\psi\in W_ 0^{1,p}(\Omega) \cap L^{\infty}(\Omega)\). The regularity results are used to prove the following assertion on homogenization: Let strongly nonlinear operators \(A_{\epsilon}\) H-converge to A, \(c_{\epsilon}\) converge to \(c_ 0\) in \(\sigma(L^{\infty}(\Omega),L^ 1(\Omega))\). Then (under some assumptions on \(A_{\epsilon},c_{\epsilon})\) solutions \(u_{\epsilon}\) of the problem \[ A_{\epsilon}u_{\epsilon} + c_{\epsilon}(x) g(x,u_{\epsilon}) = f\in H^{-1}(\Omega) \] weakly converge in \(H^ 1_ 0(\Omega)\) to the solution \(u_ 0\) of the problem \[ A_ 0u_ 0+c_ 0(x)g(x,u_ 0)=f \] and \[ A_{\epsilon} u_{\epsilon}\to A_ 0u_ 0\text{ in } H^{-1}(\Omega), \] \[ c_{\epsilon}(x)g(x,u_{\epsilon})\to c_ 0(x)g(x,u_ 0)\text{ in } D'(\Omega). \]
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    Sobolev exponent
    0 references
    regularity
    0 references
    strongly nonlinear
    0 references
    growth
    0 references
    Carathéodory function
    0 references
    variational inequalities
    0 references
    obstacle
    0 references
    homogenization
    0 references
    H-converge
    0 references