Search

A molecular Rayleigh scattering technique is developed to measure gas velocity, temperature, and density in unseeded gas flows at sampling rates up to 32 kHz. A high-power continuous-wave laser beam is focused at a point in an air flowfield, and Rayleigh scattered light is collected and fiber-optically transmitted to a Fabry-Perot interferometer for spectral analysis. Photomultiplier tubes operated in the photon-counting mode allow high-frequency sampling of the total signal level and the circular interference pattern to provide density, temperature, and velocity measurements. Mean and root-mean-square fluctuation measurements in both an electrically heated jet facility with a 10-mm-diam nozzle and in a hydrogen-combustor-heated-jet facility with a 50.8-mm-diam nozzle at NASA John H. Glenn Research Center at Lewis Field are presented. Measurements are compared with hot-wire anemometry, cold-wire thermometry, and temporally resolved particle image velocimetry to validate the Rayleigh technique.

An analytical--numerical model is developed to calculate the deformation and spreading of axisymmetric and two- dimensional liquid drops in a gas stream for small Ohnesorge numbers and large Reynolds numbers. The model is based on an approximate series solution of the Navier--Stokes equations written for the interior flow of the drop and assuming a quasi-static gas flow. The pressure distribution on the drop surface is provided using a perturbation analysis about a trivial zeroth-order solution. The axisymmetric formulation is used for calculating the deformation of the spherical droplets, and the two-dimensional formulation is used to develop a model for predicting the spreading of the cross section of liquid jets in crossflows. The penetration of the liquid jet in the gas flow is calculated, and the results are compared with experimental and empirical results.

The instability of a compressible gas jet injected into a coflowing viscous liquid medium has been studied without considering the effects of gravity. A linear temporal instability analysis is used in monitoring the wave growth rate as a function of wave number. Results show the gas Weber number, Mach number, Ohnsorge number, density ratio and velocity ratio are the parameters that control the growth rates of the instability. There was an increase in unstable wave numbers and in the growth rates of the unstable modes when the Mach number was increased.

The interaction between a normal shock wave and a boundary layer along a wall surface in internal compressible flows causes a very complicated flow. When the shock is strong enough to separate the boundary layer, the shock is bifurcated and one or more shocks appear downstream of the bifurcated shock. A series of shocks thus formed, called 'shock train', is followed by an adverse pressure gradient region, if the duct is long enough. Thus the effect of the interaction extends over a great distance. The flow is decelerated from supersonic to subsonic through the whole interaction region. In this sense, the interaction region including the shock train in it is referred to as 'pseudo-shock' in the present paper, as Crocco called it. The shock train and pseudo-shock strongly affect the performance and efficiency of various flow devices. In the present review some fundamental characteristics of the shock train and pseudo-shock are first described. Some simple predictions are made to simulate these very complicated phenomena. Pseudo-shocks appearing in various flow devices are explained. Control methods of the pseudo-shocks are also described. Finally, the current understanding of self-excited oscillation of pseudo-shock is reviewed.

Some dense gas phenomena that may be observed in nonclassical aero-gas dynamics for simple geometries are investigated by simulation and analyses of selected cases. Shock phenomena over configurations for dilute, perfect gases are compared to two-dimensional shock wave refraction phenomena for dense gases in the high pressure and density area in the thermodynamic critical point. To compute the time-dependent, two-dimensional Euler equations of a van der Waals, a two-step predictor-corrector total variation diminishing scheme is employed. The results are discussed.

It is possible to compute highly underexpanded freejets with vibrational relaxation by applying the method of characteristics (MOC) to the simulation of several steady, inviscid, supersonic flowfields. The MOC was implemented numerically to simulate a highly underexpanded freejet flow with vibrational nonequilibrium, and was also applied to freejets of gases with complete vibrational freezing or equilibrium. It was demonstrated that the computational techniques were valid for vibrationally frozen flows and showed the maximum angular spacing of the characteristics launched from the nozzle lip needed to maintain the accuracy of the solution downstream.

A 2D finite element methodology was developed for simulating two-phase dusty detonations subject to inert particles. The technique was applied to a combustion model to determine the effect of particle conditions on detonation success. Results show that as the particle diameter decreases the drag, work and heat holdings on the gas increase. Moreover, detonation failure occurs when the particle diameter became significantly small.

Experiments on the use of N(sub 2)O to increase shock tunnel test time were performed in the electric arc-driven shock tube at the NASA Ames Research Center. A total of 29 tests were conducted, 17 with air in the driven tube, 10 with N(sub 2)O/N(sub 2) in the driven tube and 1 with argon in the driven tube. Results reveal that the test times estimated from light histories increased by 60-100% when substituting N(sub 2)O/N(sub 2) for air in the driven tube and when operating at overtailored conditions.

A generalized model for thesolution of one-dimensional, steady, compressible gas flow problems is presented. The model represent each independent flow variable by a constant change coefficient. These variables are then determined numerically without resorting to the traditional iterative solution. The computational results showthat a change in the value of the coefficient affects the maximum change the flow can sustain without altering the initial flow conditions.

The mechanisms behind heat transfer in supersonic flows are presented. Results show that heat transfer peaks and their periodic distribution occur often in supersonic three-dimensional gasflows near vehicles. A heat flux measurement using thermal sensitive coatings and microcalorimeters, and the visualization of a surface flow reveal heat transfer peculiarities. These peculiarities are caused by many factors, and their variability complicates their detection and study.

The strength of streamwise vorticity generated by primary tabs can be boosted by secondary tabs, according to research designed to assess jet flow with the secondary tabs at different angles with respect to the primary tabs. The presence of the primary tabs induces a faster velocity decay than no tabs. Decay is even faster if secondary tabs are added. The streamwise vorticity dissipates within six diameters downstream of the exit plane.

A two-phase compressible flow solver along with a new Lagrangian parcel-adaptive method enables a more efficient study of particle and droplet flows. This method is used to predict a particle laden shock wave attenuation as a test case. The performance characteristics are compared with an Eulerian-Eulerian implementation, and it is found that the parcel adaptivity results in an order of magnitude savings in computational resources.

The probability of accelerating two-dimensional unsteady compressible flow computations with a generalized minimal residual (GMRES) technique employing a conjugate-gradient-like technique to accelerate the convergence of current solvers, was explored. A Newton iterative time-marching scheme that helped solve two-dimensional compressible Navier Stokes equations in a curvilinear body-filled coordinate system can be used as a preconditioner.

Experimental studies of shock modification in weakly ionized supersonic gas flows are discussed. In these experiments, a supersonic nonequilibrium plasma wind tunnel, which produces a highly nonequilibrium plasma flow with the low gas kinetic temperature at M = 2, is used. Supersonic flow is maintained at complete steady state. The flow is ionized by a high-pressure aerodynamically stabilized dc discharge in the tunnel plenum and by a transverse rf discharge in the supersonic test section. The dc discharge is primarily used for the supersonic flow visualization, whereas the rf discharge provides high electron density in the supersonic test section. High-pressure flow visualization produced by the plasma makes all features of the supersonic flow, including shocks, boundary layers, expansion waves, and wakes, clearly visible. Attached oblique shock structure on the nose of a 35-deg wedge with and without rf ionization in a M = 2 flow is studied in various nitrogen-helium mixtures. It is found that the use of the rf discharge increases the shock angle by 14 deg, from 99 to 113 deg, which corresponds to a Mach number reduction from M = 2.0 to 1.8. Time-dependent measurements of the oblique shock angle show that the time for the shock weakening by the rf plasma, as well as the shock recovery time after the plasma is turned off, is of the order of seconds. Because the flow residence time in the test section is of the order of 10 [micro]s, this result suggests a purely thermal mechanism of shock weakening due to heating of the boundary layers and the nozzle walls by the rf discharge. Gas flow temperature measurements in the test section using infrared emission spectroscopy, with carbon monoxide as a thermometric element, are consistent with the observed shock angle change. This shows that shock weakening by the plasma is a purely thermal effect. The results demonstrate the feasibility of both sustaining uniform ionization in cold supersonic nitrogen and airflows and the use of nonequilibrium plasmas for supersonic flow control. This opens a possibility for the use of transverse stable rf discharges for magnetohydrodynamic energy extraction and/or acceleration of supersonic airflows.

The properties of turbulence generated by uniform fluxes of monodisperse spherical particles moving through a uniform flowing gas were studied experimentally, emphasizing the properties of the region surrounding individual wake disturbances, i.e., the turbulent interwake region. Mean and fluctuating values, probability density functions, and energy spectra of streamwise and cross-stream velocities were measured within a counterflowing particle/air wind tunnel using particle wake discriminating laser velocimetry. Test conditions included nearly monodisperse glass spheres having diameters of 0.5-2.2 mm, particle Reynolds numbers of 106-990, mean particle spacings of 13-208 mm, particle volume fractions less than 0.003%, direct rates of dissipation of turbulence by particles less than 4%, and turbulence generation rates sufficient to yield streamwise relative turbulence intensities in the range 0.2-1.5%. The turbulent interwake region was homogeneous and nearly isotropic with probability density functions that are well approximated by Gaussian functions. Relative turbulence intensities were correlated effectively based on an analogy to the properties of isotropic grid-generated turbulence by scaling with the mean particle spacing normalized by the particle wake momentum diameter. For present turbulence generation conditions the turbulent interwake region had turbulence Reynolds numbers of 0.4-3.5 and was in the final decay period where vortical regions fill the turbulent interwake region but are sparse. This implies enhanced rates of dissipation of turbulent kinetic energy and decreasing macroscale/microscale ratios of the turbulence with increasing Reynolds numbers, as opposed to increasing ratios with increasing Reynolds numbers typical of conventional fully developed isotropic turbulence.

An optical method for generating localized, controlled perturbations has been developed for use in supersonic and hypersonic flowfields. The thermal spot disturbance is generated when the pulsed beam from a laser is focused at the desired origin and a small region of the gas is ionized. After recombination, the thermal spot persists as a region of heated gas that convects with the local flow velocity. The perturbation is approximately spherical and several millimeters in diameter, with the size dependent on the flow density. Optimal formation of the perturbation is achieved with large-F-number focusing systems. To date, the disturbance has been used as a freestream perturbation for supersonic receptivity experiments as well as forward-facing cavity and blunt-body shock oscillation studies.