Eigenvalues and bifurcation for the biharmonic operator with pointwise gradient constraint (Q1315155)

The paper is concerned with the stationary eigenvalue problem \[ (SEP)_ \rho \begin{cases} (\lambda, u)\in \mathbb{R} \times (\mathbb{K} \cap S_ \rho) \\ \langle Lu,v - u\rangle + \int_ \Omega g(x,u)(v - u)dx + \lambda \int_ \Omega u(v - u)dx \geq 0\quad \forall v \in \mathbb{K}\end{cases} \] and the evolution eigenvalue problem \[ (EEP)_ \rho \begin{cases} \Lambda: [0,T] \to \mathbb{R},\;{\mathcal U} \in W^{1,1}(0,T;L^ 2(\Omega)),\;{\mathcal U}(t) \in \mathbb{K} \cap S_ \rho\\ \langle L{\mathcal U}(t),v - {\mathcal U}(t)\rangle + \int_ \Omega g(x,{\mathcal U}(t))(v - {\mathcal U}(t))dx + \\ + \Lambda(t) \int_ \Omega {\mathcal U}(t) (v - {\mathcal U}(t))dx + \int_ \Omega {\mathcal U}'(t) (v - {\mathcal U}(t))dx \geq 0\quad\text{for a.e. }t, \forall v \in \mathbb{K} \\ {\mathcal U}(0) = u_ 0 \in \mathbb{K} \cap S_ \rho\end{cases} \] where \[ S_ \rho = \{u \in L^ 2(\Omega): \int_ \Omega u^ 2 dx = \rho^ 2\},\quad Lu = \Delta^ 2 u, \] \[ \mathbb{K} = \{u \in W^{2,2}_ 0(\Omega): | Du| \leq \varphi\}. \] If \(L\) and \(\mathbb{K}\) are substituted by \(L_ 1 = -\Delta\) and \[ \mathbb{K}_ 1 = \{u \in W_ 0^{1,2}(\Omega): \psi_ 1 \leq u \leq \psi_ 2\}, \] problems \((SEP)_ \rho\) and \((EEP)_ \rho\) have been treated by \textit{G. Chobanov}, the second author, and \textit{D. Scolozzi} [Rend. Accad. Naz. Sci. Detta XL, V. Ser. Mem. Mat. 14, No. 1, 139-162 (1990; Zbl 0729.35088), and Nonlinear Anal., Theory Methods Appl. 15, No. 3, 199-215 (1990; Zbl 0716.49009)]. Under suitable assumptions, it has been shown that \((EEP)_ \rho\) has a unique local solution and \((SEP)_ \rho\) has one or even a sequence \((\lambda_ h,u_ h)\) of solutions with \(\lambda_ h \to -\infty\), provided that a symmetry condition is satisfied. Moreover, if \(g(x,0) = 0\) and \(0 \in \mathbb{K}_ 1\), the pairs \((\lambda_ 0,0)\) which accumulate solutions \((\lambda_ \rho,u_ \rho)\) of \((SEP)_ \rho\) with \(u_ \rho \neq 0\) (bifurcation points), have been studied by the first two authors in Atti Accad. Naz. Lincei, VIII. Ser., Rend., Cl. Sci. Fis. Mat. Nat. 81, No. 3, 259-269 (1987; Zbl 0671.58029)] and by the first author in Ann. Fac. Sci. Toulouse, V. Sér., Math. 10, No. 2, 215-258 (1989; Zbl 0694.58042) and Ann. Fac. Sci. Toulouse, V. Sér., Math. 11, No. 1, 39-66 (1990; Zbl 0717.49010). Following the same theoretical approach, the case of the operator \(L\) and the constraint \(\mathbb{K}\) is now considered. Analogous results are announced for \((SEP)_ \rho\), \((EEP)_ \rho\) and the bifurcation problem. Particular relevance is given to the ``non-tangency condition'' between \(\mathbb{K}\) an \(S_ \rho\), whose features are quite different with respect to the case of \(L_ 1\) and \(\mathbb{K}_ 1\).