Publications

The extraction of the modal parameters for closely spaced modes in the frequency domain is a common problem. However, it is made more difficult if the damping for the closely spaced modes is high. Data from a structure with more than three percent viscous damping is presented which exhibits this phenomenon. Traditional experimental techniques failed to identify all the modal parameters of three closely spaced modes. Mode shapes from an analytical model are manipulated to produce a modal filter which is used to calculate enhanced frequency response functions from which the modal parameters can be more readily identified. Discussion of the advantages and disadvantages of this technique as compared with traditional frequency response function enhancement techniques will be presented.

New Mexico State University organized an effort to perform static and dynamic damage-detection tests on the Interstate-40 bridge over the Rio Grande at Albuquerque. The opportunity was available because the 425-ft-long bridge was soon to be replaced. Sandia National Laboratories was asked to provide and operate a shaker that could exert 1000-lb peak amplitude forces for both sinusoidal and random excitations between 2 and 20 Hz. Two Sandia departments collaborated to design and build the shaker, using existing major components connected with Sandia-designed and -fabricated hardware. The shaker was installed and operated successfully for a series of five modal and sinusoidal response tests.

A technique to localize errors between two modal models is presented. Mode shape difference are calculated from each model. A global comparison of the ratios of these corresponding differences is used to identify the physical locations on the structure where stiffness differences exist between the two models. Some of the strengths and limitations of the technique are illustrated using the mode shapes of two similar finite element models with a known stiffness difference. The technique is then applied to a two link robot arm for which a finite element model exists and a modal test has been conducted. The results of the error localization aid the selection of physical parameters to be updated in the finite element model. Sensitivity methods are used to correlate the finite element model to the modal test. The results of the correlation are presented. 5 refs., 3 figs., 3 tabs.