244 T.J. Beberniss et al. Fixture Panel LED panel lights Halogen lamps for panel heating Static 3D-DIC Cameras FLIR Camera Fig. 22.9 Laboratory testing arrangement for static buckling test [7] then raised to 10ı. Observe in Fig. 22.10(a) the marked difference in the dynamic displacement due to loading changes caused by the impinging shock. In Fig. 22.10(b) the displacement power spectral density (PSD) exhibits all the signs of a hardening geometric nonlinearity—the positive shift in frequency and the broadening of the response peaks in the frequency domain. The frequency broadening stems from the stochastic (turbulent boundary layer) input to a structural system where stiffness is a function of displacement. Finally, in Fig. 22.10(c, d), images of the surface pressure (PSP) for the no-shock/shock cases are included to show the spatially disparate pressure loading. These PSP images were captured by looking through a window in the tunnel sidewall, thus the skewed perspective. In the second year of testing a window was added to the bottom tunnel wall and the images were much less skewed. Next, consider a case (Fig. 22.11) from the second year of testing, where the panel response to SBLI and heated flow conditions was studied using a fixed 8ı shock generator. In Fig. 22.11(a), the panel center displacement time history is shown for both heated and unheated flow conditions. The corresponding displacement PSD is displayed in Fig. 22.11(b). The heated wind tunnel flow quickly overwhelmed the PSP and so the surface pressure image of Fig. 22.11(c) is for the unheated case only. Interestingly, the panel center displacement PSD of the post-buckled panel does not differ appreciably from the unheated/non-buckled case. In the heated flow condition, the panel was buckled 2 mm (approximately 3 panel thicknesses) into the flow; however, the dynamic response is clearly not as great as was exhibited in previous experiments. One new variable in the upcoming experiments will be the ability to adjust the cavity back pressure to accentuate/tune the panel dynamic response. Ideally, the conditions leading to dynamic snap-through will also be identified. To that end, new external sources of heating are being explored to (1) exert greater control over the magnitude of panel buckling, and (2) continue to use the PSP (and TSP) during the experiments. The results of this early experimentation can be seen in Fig. 22.12, which displays (a) the full-field temperature field and (b) the state of the panel (displacement) from nearly 80ıC to room temperature. Full-field DIC images of the panel displacement field correspond to several discrete temperature points are also denoted in Fig. 22.12(b). These results were obtained by first heating the panel using the halogen lamps and associated testing arrangement shown in Fig. 22.9, and then taking DIC and FLIR images as the panel cooled to room temperature. The panel is initially (nominally) flat but the effect of imperfections can be observed by considering the panel center displacement (black curve), of Fig. 22.12(b). In contrast to a perfectly-flat panel reaching a critical point beyond which symmetric buckling occurs, the imperfect panel deflects with temperature beyond the critical point until another stable, asymmetric equilibrium is realized. A good discussion of this very issue for panel buckling/acoustic experiments is discussed by Murphy [13]. At maximum temperature (approx. 80 ıC) the panel did exhibit multiple (asymmetric) stable equilibria during the experimentation and so had reached the point where the secondary branch, described by Murphy, had appeared.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==