On the optimal column shape (Q1385902)

The classical problem of the shape of a column that is the body of revolution of a plane curve and can support the maximal axial loading is one of the most important problems of elasticity theory. Assume that the length and volume of the column are given and its ends are rigidly fixed. The solution to this problem was suggested, for the first time, by Keller and Tadjbakhsh in 1962. However, their deduction of the necessary conditions of optimality, from which the solution was obtained, is based on the implicit (and incorrect!) assumption that ultimate admissible load is a simple eigenvalue of the corresponding differential operator. Olhoff and Rasmussen in 1977 and Cox and Overton in 1992 tried to prove that the result of Keller and Tadjbakhsh is not true. The purpose of this paper is to show that, in spite of some errors in the proofs, the final result of Keller and Tadjbakhsh is true. Moreover, the authors prove a theorem of existence of the optimal column.