PSI - Issue 31

Alok Gupta et al. / Procedia Structural Integrity 31 (2021) 15–21 Alok Gupta et al. / Structural Integrity Procedia 00 (2019) 000–000

18

4

refer Fig. 4) were mounted on the dummy block to measure the response of the dummy block/bracket. A tri-axial accelerometer (in the axial, radial and tangential directions) and a uniaxial accelerometer (in the axial direction) were mounted on the fixture to control the input to the bracket. Firstly, a resonance search test was carried out to establish the fundamental vibration mode of the bracket assembly. This test was performed with the axial response of the bracket controlled at ‘25g’ acceleration level (maximum). This control loop was chosen to ensure that the bracket response would not go beyond ‘25g’ level, meaning that the control loop would automatically drop the input level down at and around the resonance point. Subsequently, a series of dwell tests were carried out to establish the ‘g’ capability of the bracket. In a dwell test the bracket was excited at its fundamental frequency at a predefined and controlling response ‘g’ level of the dummy block. At first a trial dwell test was performed at 84Hz frequency and ‘30g’ response level to ensure that the bracket assembly, the sensors, the shaker and the data acquisition system performed as expected. The trial test was stopped after 2050 s (i.e. at 172,239 cycles) and the mounting bolts were checked for tightness at the specified torque values. Subsequently, the dwell tests, starting with ‘40g’ response level, were performed to establish the brackets ‘g’ capability. Each dwell test was defined as a test run in which the response ‘g’ level at the dummy block was kept constant. The response ‘g’ level in each subsequent test run was increased to the next planned level, which in this case were the ‘60g’ and then ‘80g’ levels, until a sequential failure of the bracket struts/connectors were observed. The last test run was with a low ‘20g’ response level to limit the chances of any possible damage to the surrounding equipment and sensors.

Axial deformation mode, 85.6 Hz

Axial

Tangential

Radial

Fig. 3. FE predicted fundamental vibration mode at 85.6Hz depicting an axial deformation mode of the bracket assembly. Axial Direction

Excitation Direction

Fig. 4. Test setup - bracket assembly and fixture mounted on the shaker table with instrumentation

3. Test Results and Discussion From the resonance search (frequency sweep) test, the Frequency Response Curve (FRC) in the axial and tangential directions (refer Fig. 4) are shown in Fig. 5a and Fig. 5b respectively. The fundamental vibration mode was observed at 84 Hz (compared to 85.6 Hz from FE); a dominant axial bending mode of the bracket. The amplification observed was 73 meaning a low damping ratio of 0.68% for the fundamental mode of the bracket assembly. The 2nd less responsive vibration mode was at 183 Hz which is the tangential mode of the bracket assembly. Since the excitation given was in the axial direction and the response was measured in the tangential direction, the inference of damping from the response amplification for the second mode could not be justified. The dwell test runs used with their corresponding ‘g’ response levels and accrued number of cycles until failure of the bracket are listed in Table 2. The marked sequence of failure (Locations 1-5) in the bracket is indicated in Fig. 6. Based on the observations presented in Table 2, the bracket survived for 10 6 cycles at ‘60g’ acceleration level without any failure of the struts or connectors. However, during the test run at ‘60g’level, an increase in input excitation level was observed after 635,866 cycles. The test was stopped at this point and a resonance search test was carried out which indicated a slight drop in fundamental modal frequency of the assembly (of 3 Hz). A careful visual examination at this point did not show any signs of crack initiation or a crack in any of the struts or connectors. The dwell test was continued at 81 Hz frequency to complete 1x10 6 cycles at ‘60g’ level. Upon completion of this test run, a detailed inspection was carried out which again did not reveal any cracks or fractures in any of the struts or connectors.