Morse theory for \(C^ 1\)-functionals and Conley blocks (Q1904453)

The authors consider an open subset \(\Omega\) of a Hilbert manifold, a functional \(f\) in a neighborhood of \(\overline {\Omega}\) and the set \(K_f\) of critical points of \(f\) in \(\overline {\Omega}\). In the following \(\lambda\) is the notation for the variable in the diverse polynomials. They introduce an index \(i_\lambda\) generalizing the Morse polynomial \(m_\lambda\) for the \(C^2\)-case and nondegenerate critical points of \(f\) to the following cases: (1) the critical points of \(f\) may be degenerate; (2) \(f\) and/or \(\Lambda\) are of class \(C^1\). Furthermore (3) \(\overline \Omega\) need not be a strip \(f^b_a = \{x \in \Lambda;\;a \leq f(x) \leq b\}\). Moreover they define a class of closed sets (called Conley blocks) and an index \(I_\lambda\) of Conley blocks. \(i_\lambda\) is a differential invariant, \(I_\lambda\) a topological invariant. They prove generalized Morse relations for these indices. Often one can not assume a priori that the critical points are nondegenerate and functionals that are used in partial differential equations are often \(C^1\) but not \(C^2\). The authors give a typical application of this theory. \(I_\lambda\) is inspired from the Conley index [\textit{C. Conley}, `Isolated invariant sets and the Morse index', Reg. Conf. Ser. Math. 38 (1978; Zbl 0397.34056)]. The generalized Morse index was announced by the authors in C. R. Acad. Sci., Paris, Ser. I 315, No. 8, 883-888 (1992; Zbl 0768.58010).