The first component of the fixture is a large circular plate with a diameter of 460mm, and a thickness of 20mm. The second component of the fixture is a smaller square plate with a 180mm height and width and a 20mm thickness. The square plate also contains three spherical standoffs of 20mm height that localize the contact area between the two plates. Both plates were machined from 4140 Alloy Steel. The fixture is assembled by bolting the two components together with M-16 bolts and load washers to measure the static and dynamic preload in the bolts. The fixture was used to study the behavior of the bolted joint due to impulsive loading. In the problem being simulated by the fixture, the load path passes through the circular plate and into the square plate. The primary goal of the experiments that were conducted was to determine how preload in the bolts affects the force transmission across the interface. First, modal analysis was conducted on the test fixture for various preload levels and impact amplitudes, and then high amplitude impacts were applied to the test fixture with a Hopkinson bar in order to approach a more realistic loading scenario. The results of the modal impact tests were used to demonstrate the effect of bolt preload on the linear response, while the results of the high amplitude impact testing were used to demonstrate that the force amplification characteristics across the boundary can be explained with a lumped parameter two degree of freedom dynamic model. According to work by Peairs, Park, and Inman (2001), bolt preload level is important in a bolted interface because “varying the preload tension in a bolt changes the integrity of the particular joint and also changes the structure's global dynamic properties.” It was expected that these changes would be reflected in the mode shapes, and that the experiments performed could quantify the particular properties of the test fixture that are most affected by changes in preload. Change in modal frequencies due to varying bolt preload is also examined in literature as described in the work by Chang, Erickson, Lee, and Todd (2004), where an experiment was presented as part of a damage detection study that studied the percentage change of natural frequencies for five modes of vibration as a function of bolt preload level. The experimental setup consisted of two thin (¼ inch thick) metal beams affixed with a single bolt. The result of this study showed that, although the change in frequency was different for each mode, there was certainly an upward shift in natural frequency that occurred as a result of increasing the preload of the bolt. It was determined that “higher frequencies are [sic] generally a better indicator of damage because the change in frequency at low frequency ranges is very small, even though the percentage change may be slightly larger.” The nonlinear effects of bolted joints were also studied by Ibrahim and Pettit (2005), where the filtering properties of threaded interfaces are characterized by the use of transmissibility functions. Finally, in a paper from IMAC XXVIII, Adams et. al. (2010), also used transmissibility functions to investigate the effect of threaded joint preload on structural response for a particular cylindrical structure that can undergo shock loading. The substructure being studied in this paper is a component from the structure studied in Adams et. al. (2010). EXPERIMENTAL APPROACH The modal impact testing of the bolted interface between the square and circular plate was carried out based on the results of a finite element analysis of the system that produced the estimated real normal mode shapes and undamped natural frequencies. The analysis showed that a total of 46 impact testing points would be sufficient to observe modes of vibration up to 6 kHz. The 46 points were distributed with 9 points on the square plate, 31 points on the circular plate, and 6 points near the accelerometers for driving point measurements. The modal grid that was used is shown in Fig. 2. The equipment that was used for the test consisted of six PCB 356A32 100 mV/g tri-axial accelerometers rigidly attached with superglue, three RS Technologies Bolt Load Force Transducers (Model 054216) connected to PCB Strain Gauge Signal Conditioners with digital display units (PCB Series 8159), two PCB modal impact hammers (Model 086C03 and Model 086C01), and a VXI Mainframe rack unit (Agilent E8408A) with three 4-16 channel 51.2 KSa/s A/D+DSP circuit cards (Agilent E8491B, E1432A, and HP E1432A). The accelerometers were placed at each end of the three bolts in the joint. The test setup is shown in Fig. 3. Fig. 2 Modal Grid 568
RkJQdWJsaXNoZXIy MTMzNzEzMQ==