Hotwire experiments were performed to assess the effects of suspended 100 nm diameter aluminum oxide (Al2O3) nanoparticles in air on the rate of convective heat transfer. The particle mass loading (0.01% < SL < 0.1%) was varied over a range of Reynolds numbers (1420 < Re < 10,310), while monitoring the global Nusselt number. The results indicate up to a 45% increase in the Nusselt number over the baseline air case for a particle mass loading of 0.1% at a Reynolds number of 7,500. The enhancement to heat transfer is an order of magnitude larger compared to results from related heat transfer experiments reported in the literature that used micron-sized aerosols at similar particle mass loadings. Potential application of the dispersion of nanoparticles into the intake air of a high-speed propulsion system is discussed. Nomenclature

Shock wave formation and acceleration in a high-aspect ratio cross section shock tube were studied experimentally and numerically. The relative importance of geometric effects and diaphragm opening time on shock formation are assessed. The diaphragm opening time was controlled through the use of slit-type (fast opening time) and petal-type (slow opening time) diaphragms. A novel method of fabricating the petal-type diaphragms, which results in a consistent burst pressure and symmetric opening without fragmentation, is presented. High-speed schlieren photography was used to visualize the unsteady propagation of the lead shock wave and trailing gas dynamic structures. Surfacemounted pressure sensors were used to capture the spatial and temporal development of the pressure field. Unsteady Reynolds-Averaged Navier-Stokes simulation predictions using the shear-stress-transport turbulence model are compared to the experimental data. Simulation results are used to explain the presence of high-frequency pressure oscillations observed experimentally in the driver section as well as the cause of the initial acceleration and subsequent rapid decay of shock velocity measured along the top and bottom channel surfaces. A one-dimensional theoretical model predicting the effect of the finite opening time of the diaphragm on the rate of driver depressurization and shock acceleration is proposed. The model removes the large amount of empiricism that accompanies existing models published in the literature. Model accuracy is assessed through comparCommunicated by isons with experiments and simulations. Limitations of and potential improvements in the model are discussed.

No abstract