15 Investigation of a Novel Pseudo Ambient Vibration Testing Approach 139 the writers believed that an experimental evaluation of the testing approach would incorporate many of the uncertainties and challenges associated with dynamic testing of full-scale structures more realistically than would be possible from a purely analytical study of this approach. The experimental testing program devised for this study consisted of a comparative study of the modal parameters identified for a large-scale steel structure by conventional ambient vibration testing with no artificial dynamic excitation applied to the structure, and by the proposed pseudo ambient testing approach in which uncorrelated white noise excitation was applied to the structure in a spatially distributed manner by a network of low-cost and small scale shakers. The pseudo ambient excitation cases considered for this investigation included variations in the quantity and spatial locations of the dynamic inputs to the structure and to the characteristics of the dynamic excitation. Although the full testing program included a total of 10 different excitation cases, the focus of this paper is on the results obtained from a subset of four of these test cases. These test cases discussed in this paper include an uncontrolled ambient excitation case and three cases in which pseudo ambient excitation having different characteristics was applied to the grid by a network of 15 shakers. The vibration responses were measured and processed in the same way for all test cases. The dynamic excitation applied to the structure in each test case was also considered to be unmeasured and uncontrolled during the modal parameter identification stage. The details of the test structure and the experimental program are described in the following sections. 15.2.1 Test Structure A large steel structure was used as the test specimen for this research. The model can be nominally described as a grid structure, although it was designed and constructed to notionally represent various behavior characteristics common to many different types of civil infrastructure systems. The grid structure is simply supported with a single span length of 7,315.2 mm and a width of 2,743.2 mm. The grid framing system consists of three continuous longitudinal girder lines spaced on 1,371.3 mm centers and 7 transverse girder lines spaced on 1,219.2 mm centers. The ends of each longitudinal girder are supported by 1,066.8 mm tall steel pedestals and pin and roller bearings. The steel pedestals sit on the approximately 130 mm thick unreinforced concrete slab on grade floor in the laboratory. The 3 longitudinal girder sections and the 14 transverse girder sections are all W8 10 rolled steel beams that are joined together with bolted connections employing top and bottom cover plates and vertical clip angles which results in fairly rigid connections between the members. A photograph of the test structure and a schematic illustrating the grid numbering convention employed in this study are provided in Fig. 15.1. 15.2.2 Test Setup and Measurement Equipment The grid structure was instrumented for the vibration testing using a total of 21 accelerometers (Model 393C from PCB Piezotronics, Inc.) placed at the intersections of the longitudinal gridlines 1–3 and the transverse gridlines B–F, H and J as defined in Fig. 15.1. The accelerometers were attached to the structure using magnets and were oriented to measure the vertical accelerations of the structure. A total of 15 shakers (tactile transducers) were attached to the grid model using c-clamps to provide the controlled pseudo ambient dynamic excitations in the vertical direction. The shakers were installed at the intersections of longitudinal gridlines 1–3 and transverse gridlines B, C, D, E and F. Since the weight of the 15 shakers could potentially alter the modal characteristics of the structure, they were left in place on the structure during all test cases. The vibration measurements were recorded using National Instruments Model 9234 dynamic signal acquisition modules. Uncorrelated white noise excitation signals specific characteristics were generated and sent to the shaker amplifiers through independent channels of a National Instruments Model 9264 analog output module. The time duration of the vibration measurements recorded in each test case was 60 min. The digitized measurements were subsampled to an effective sampling rate of 413.05 Hz in before being saved for subsequent data analysis. 15.3 Vibration Test Cases and Dynamic Excitation The results obtained from a total of four different vibration test cases are presented and discussed in this paper. Case 1, represents the baseline ambient vibration test for the structure. The only dynamic excitation acting on the structure was from uncontrolled and unmeasured ambient sources present in the laboratory facility. This test case is representative of the ambient excitation that might act on a bridge that is closed to traffic. The ambient dynamic excitation of the grid structure in this case
RkJQdWJsaXNoZXIy MTMzNzEzMQ==