22 Nonlinear Response of a Thin Panel in a Multi-Discipline Environment: Part I—Experimental Results 241 Fig. 22.6 RC-19 tunnel test-section [10] The flexible panel specimen was designed so that an appreciable number of the structural dynamic modes would be below 1000 Hz, just like a traditional aircraft panel. To accommodate this test article arrangement, the RC-19 top tunnel test section was replaced with three-sections. The first section, containing pressure ports, was located upstream of the test article. The second section included the compliant panel. The last section connected the modified test section to the tunnel exhaust section and, importantly, was ported to a cavity with the top window (see Fig. 22.6) behind the compliant panel. These pressure ports were added to the top section of the tunnel wall downstream of the compliant panel to equalize the pressure between the panel and the top window. Equalization of the pressure was necessary to prevent yielding of the panel during tunnel start-up. The top window allowed for the use of 3D Digital Image Correlation (DIC) to obtain the full field displacements of the panel during testing. The sides of the tunnel also contained large windows to allow for pressure sensitive paint (PSP) illumination and implementation of a shadowgraph set-up to visualize the flow over the test panel. During testing, and to characterize the fluctuating pressure on the compliant panel, a fast reacting pressure sensitive paint was applied to the flow side. The experimental procedure, described in greater detail in [11], began as follows. First the exhaust section was used to reduce the tunnel pressure to approximately 0.2 atm, and then the tunnel was started by allowing an inflow of air and setting the inlet total pressure to 2.65 atm. This resulted in a static pressure on the panel of 0.34 atm with the shock generator at 0ı. The inflowing air was generally unheated thus creating a static temperature of 112 ıC and panel equilibrium temperature of 0.5 ıC. The frame and panel temperatures were monitored at single locations each on the compliant panel and frame throughout the test using K-type thermocouples. Test measurements did not begin until the differential temperature between the frame and test panel, initially at room temperature, reached 6.5 ıC. The temperature difference resulted in a tensile preload on the panel, slightly altering the dynamic response. Once a stable temperature difference was achieved, the DIC cameras were triggered and images were recorded for 20.8 s at a sampling frequency of 5 kHz. The PSP was recorded for 5 s at 500 Hz. At the same time, the laser vibrometry data and strain gage data were recorded on a separate data acquisition system for 60 s at a sampling rate of 10 kHz. The shock generator angle was increased in 2ı increments, and the process was repeated until the shock generator reached 10ı. The boundary layer thickness without the shock generator is approximately 7.6 mm where the flow first meets the panel and grows to 10.2 mm at the end of the panel. Some changes were necessary for the second year of testing. First, the frame and panel were machined from a single block of ANSI 4150 alloy steel. A panel/frame bond-line failure during the first year of test spurred this redesign. This time the integral frame/test panel was machined from a steel block leaving the same frame/specimen dimensions as in the first year of testing. At the same time, a rigid 12.7 mm control specimen was also procured. A window was added to the bottom wall test section to allow for viewing the PSP on the flow-side surface of the compliant panel. One other major change to the testing set-up was the move from a variable angle shock generator to a fixed, 8ı angled wedge. The shock generator was placed in the bottom wall of the tunnel turning the flow, resulting in an oblique shock-wave angle of 39ı emanating from the tunnel bottom wall. The shock generator could translate 170 mm in the flow direction allowing the shock to impinge from the compliant panel leading edge to near the panel mid-point as shown in Fig. 22.6. Additional lighting was added to the test so that the high-speed camera dedicated to the PSP measurement could sample at a higher frequency. This improvement allowed for both longer PSP measurement and a greater sampling frequency. Specifically, the cameras used for the DIC and
RkJQdWJsaXNoZXIy MTMzNzEzMQ==