POST-PROCESSING AMBIENT AND FORCED RESPONSE BRIDGE DATA TO OBTAIN MODAL PARAMETERS

In this paper, post-processing methods were applied to three different sets of multiple channel time domain vibration response data taken from the Z24 highway bridge in Switzerland. The objective of this exercise was to compare the mode shapes resulting from each of these data sets with one another. Ideally, all of the tests should yield the same mode shapes, although perhaps not all modes would be excited in each case. The three different test cases were; Case 1: Two shaker response data, including the simultaneously measured excitation forces. The shakers were driven by uncorrelated random signals. Case 2: Impact response data, including three reference responses but no measured forces. The impact force was provided by a 100 kg. drop weight impactor. Case 3: Ambient response data, including three reference (fixed) responses. Excitation was provided by automobile traffic on an adjoining bridge. Excitation forces were measured in Case 1, so multiple reference Frequency Response Functions (FRFs) were calculated. Since no forces were measured in Cases 2 & 3, a different set of measurements called ODS FRFs ([1], [2]) were calculated for these cases. All sets of FRFs and ODS FRFs were then "curve fit" to obtain estimates of the modal parameters of the bridge. Modal frequencies are listed in tables for comparison, and the Modal Assurance Criterion (MAC) was used to compare mode shapes.