Fig. 6 Frequency Response Functions for 20000N Preload Level A second experiment was conducted with a Hopkinson bar as discussed previously. The primary purpose of this experiment was to impact the fixture with much higher-level impacts and investigate the system behavior. Since the forces used to impact the fixture were only characterized by the pressure used to launch the striking projectile, the actual impact force is unknown. Analysis is underway to predict these forces using a model that was created with a modal impact hammer after testing was completed. The hammer was used to strike the end of the transfer bar where the load would normally initiate. This procedure allowed for the measurement of frequency response functions. These functions will eventually be used to generate force estimates for the Hopkinson bar tests, but issues with the boundary conditions and system nonlinearities have complicated the process. Early estimates predict that the largest impacts generated in this test were on the order of 8000lbf, which is more than twenty times larger than previously generated forces. Because these forces are not completely known at this time, the analysis of the data has focused for now on the acceleration response measured on the plates. The most notable result for these large level impacts is the effect of preload on the amplification or attenuation of forces across the interface. While the impact was always applied to the center of the circular plate in order to avoid effects due to asymmetrical loading, the acceleration response was sometimes larger on the smaller, square plate. The acceleration response near the end of the measurement window for a 20 psi impact and a hand tight bolt preload condition is shown in Fig. 7. The responses for the 50ft-lb and 100ft-lb preload level are shown in Fig. 8 and Fig. 9, respectively. Each plot shows the acceleration response for accelerometers located directly across from one another, with one on each plate. These accelerometers were located at either end of the bolted joint. For the hand tight preload level, the bolts were tightened to a snug level with no tools. This procedure resulted in low coupling between the two plates and a large amount of relative motion. In this case, the acceleration response on the circular plate was larger. When the bolts were tightened to 50ft-lb, the response changed drastically. The responses were noticeably more synchronous as the coupling in the system was increased. For this case, the smaller square plate actually responded at a higher level than the larger plate, where the impact occurred. The results for the 100ft-lb preload followed the same trend, with even a greater amplification of force across the interface. This result was significant since the square plate is an analog for a sensor mount. This acceleration amplification effect could result in damage to the sensors that are used in the application of interest. 0 500 1000 1500 2000 2500 3000 3500 4000 10-4 10-3 10-2 10-1 100 101 102 103 Frequency (Hz) |FRF| 10-20 lbf Impact 200-300 lbf Impact 573
RkJQdWJsaXNoZXIy MTMzNzEzMQ==