Focal Decomposition and Some Results of S. Bernstein on the 2-Point Boundary Value Problem (Q4487491)

This well written paper deals with the 2-point boundary value problem for the second-order ordinary differential equation \[ \ddot {x}=f(t,x,\dot{x}),\quad x(t_{1})=x_{1},\quad x(t_{2})=x_{2}. \] The authors look for the number of solutions to this problem, if any, and investigate how this number varies with the endpoints \((t_{1}, x_{1})\) and \((t_{2}, x_{2}).\) The method they use relies on the Euler equation obtained from a functional of the form \(\int_{t_{1}}^{t_{2}}L(t,x,\dot {x})dt,\) where \(L\) is an appropriate function. The paper is organized as follows: Some definitions and the pendulum example are given in section 2, while in section 3 some relative old results due to S. Bernstein are presented. In section 4 the main results are formulated among which one the most important the following theorem is: Sufficient conditions are provided under which the space of all points of the form \((t_{1}, x_{1}, t_{2}, x_{2})\) is decomposed into a sequence of subsets \(\Sigma_{i},\) denoting the set of all such points for which the original problem admits \(i\) solutions. Some linear or topological properties of these sets are also described. Furthermore, by using a functional of the form \(L={{\dot {x}^{2}}\over 2}-V(x)\) the authors are able to guarantee that \(\Sigma _{i}\) is not the empty set for infinitely many indices. The paper continuous with the proofs of the results in section 6 and it closes with an appendix devoted mainly to the regularity of the Euler equation of the functional \(L.\)