UBC Theses and Dissertations
Nonlinear response of structures in regular and random waves Lipsett, Arthur William
The problem of the dynamics of a flexible offshore structure in either a regular or random sea is considered in this thesis. A simple single degree of freedom model of the structure is assumed and the relative velocity formulation of the Morison equation is used to describe the fluid force. The resulting equation of motion is a nonlinear ordinary differential equation with either harmonic or stochastic forcing depending on the wave description. Solutions are obtained for regular deterministic waves by numerical integration, various linearization methods and a new perturbation method developed in this thesis. The numerical solution is used to assess the accuracy of each of the approximate solution methods. Of these, the perturbation method is found to give the best approximation to the numerical solution over the complete frequency range of interest. For random seas the response spectrum and the mean square response are obtained by various linearization methods, the method of equivalent linearization, and by the new perturbation method. The perturbation method and the method of equivalent linearization are very similar in that they both yield the same values of effective damping. Comparison of the results obtained by a numerical simulation method with the results of the perturbation method and the widely used method of equivalent linearization shows that the perturbation method gives a better estimate of the response mean square value than does the method of equivalent linearization. For all of the approximate solution methods that are discussed it was found that the use of Hermite polynomials to represent the solution is very effective in obtaining various expected values required in the computational procedure. In addition to the average response statistics, such as the response mean square value, the probability density of the response is also considered. It is well known that the response of a linear system to Gaussian forcing is itself Gaussian. The wave force given by the Morison equation is non-Gaussian and therefore the response is also non-Gaussian but of unknown form. The hypothesis that for a linear equation, the probability density of the response is of the same form as the probability density of forcing, even for the case of non-Gaussian forcing, is investigated and verified using the results of numerical simulations. Design considerations of interest which follow from the response probability density are also discussed.
Item Citations and Data