Applications Of Sensitivity Data To Eigensolution Reanalysis Of Modified Structures

Abstract

Methods of computing eigensolution sensitivity have been known for a long time. Both exact and approximate methods are available in the literature. While eigenvalue sensitivity is used routinely in structural optimization with eigenvalue constraints, few application of eigenvector derivatives are reported in the literature. The objective of this thesis is to present an effective eigensolution reanalysis approach using first and second order eigenvector sensitivity data. In the proposed new approach, modes shapes and their derivatives are used as basis vector for eigensolution of the modified system. Comparison of numerical results with several existing eigensolution reanalysis methods shows the proposed algorithms are very effective. This will potentially make optimization using reanalysis techniques faster and more reliable.