Reliability approach to the random imperfection sensitivity of columns (Q760250)

Buckling of stochastically imperfect finite columns on a mixed quadratic- cubic elastic foundation is studied. The imperfection function is assumed to be a normally distributed random function of the space coordinate with given mean function and autocorrelation function. The problem is solved by the Monte-Carlo method. The Fourier coefficients of the initial imperfection function, expanded in terms of the buckling modes of the associated perfect column, are simulated. For each trial the buckling load is obtained numerically, and the results are used in constructing the reliability function at a specified load (defined as the probability of the buckling load exceeding this specified load). The study is a sequel to an earlier work of the author concerning the reliability of stochastically imperfect columns on a purely cubic nonlinear elastic foundation [J. Appl. Mech. 46, 411-419 (1979; Zbl 0405.73035)].