| Experimental modal analysis has become an increasingly important engineering tool during the past forty years in the aerospace, automotive, and machine tool industries. The research represented by this dissertation is concerned with developing methods which will reduce the variance of the estimates of the modal parameters determined by the frequency response function experimental modal analysis method. The primary thrust of this research is to utilize the redundant information contained in the frequency response function matrix in order to formulate the best least squares estimate of the modal parameter as well as providing additional confidence factors related to the estimation process.; Several new concepts, utilizing the redundant information in the frequency response function matrix, were formulated during this research. The concept of enhanced frequency response function is developed to use a row or column of the frequency response function matrix in order to improve frequency and damping estimates. The concepts of modal scale factor and modal assurance criterion are developed to evaluate and utilize redundant modal vector estimates in order to improve the modal vector estimates.; Finally, in order to obtain multiple columns of the frequency response function matrix simultaneously, the multiple input frequency response function method is investigated. The primary reason for this new approach to the estimation of frequency response functions is to improve the consistency between elements of the frequency response function matrix by distributing the energy of excitation throughout the system. This tends to desensitize the nonlinear effects which are a function of excitation location and response amplitude for a given level of input energy.; The concepts and techniques, evaluated in the course of this research, were demonstrated experimentally on several test objects, including a T-plate test structure, a Schweitzer sailplane, and an automotive frame. |
PDF Full Text Request