SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Two-dimensional, spatially growing, turbulent mixing layers are simulated numerically by a vortex method and the results are compared with those determined experimentally. The effects of artificial forcing on flow development are also studied. Many of the flow features which have been observed experimentally are reproduced, and good quantitative agreements between experiments and computations are obtained.
This paper presents a computational and experimental investigation of supersonic flow past a cavity in a flat plate. The source of the particular interest in this problem is the ongoing study of the aerodynamic interference effects between a separating store and its bay in the parent body. An upwind relaxation scheme, utilizing flux vector splitting and line-Gauss-Seidel iterations, is used to solve Reynolds-averaged Navier-Stokes equations. Spatial discretizations of this two-dimensional analysis are based on implicit and finite-volume methods. Turbulence is modeled and shocks are captured. The flowfield of the symmetry plane at the half-width is computationally visualized and all flow properties are computed. Experimental tests are conducted in the Langley Unitary Plan Wind Tunnel to measure wall pressures and to capture schlieren photographs. Qualitative as well as quantitative data of computations and experiments agree very well. These two vehicles of investigation are merged to show open, closed and transitional cavity flow behaviors.
An investigation of downstream boundary effects on the frequency of self-excited oscillations in two-dimensional, separated transonic diffuser flows has been conducted numerically by solving the compressible, Reynolds-averaged, thin-layer Navier-Stokes equation with a two-equation turbulence model. It was found that the unsteady diffuser flowfields are very sensitive to the location of the downstream boundary. Extension of the diffuser downstream boundary significantly reduces the frequency and amplitude of oscillations for pressure, velocity and shock. Computational results suggest that the mechanism causing the self-excited oscillation changes from viscous convective wave dominated oscillations to inviscid acoustic wave dominated oscillations when the location of downstream boundary varies from 8.66 to 134.7 throat height. The existence of a suction slot in the experimental setup obscures the physical downstream boundary and, therefore, presents a difficulty for quantitative comparisons between computation and experiment.
Two-dimensional, spatially growing, turbulent mixing layers are simulated numerically by a vortex method and the results are compared with those determined experimentally. The effects of artificial forcing on flow development are also studied. Many of the flow features which have been observed experimentally are reproduced, and good quantitative agreements between experiments and computations are obtained.
As a part of an ongoing research in aircraft icing, the leading edge separation bubble on a NACA 0012 model with a 5 min simulated glaze ice was investigated. The flow visualization methods using oil, tuft, splitter plate, smoke, and liquid crystals were employed to get reattachment line data for the leading edge separation bubble on both surfaces of the airfoil. On the upper surface, the separation bubble was found to grow larger with increasing angles of attack and reduce in size with decreasing angle of attack. On the lower surface, the bubble was found to grow larger with increasing negative angles of attack and reduce in size with increasing angles of attack. The separated flow fails to reattach beyond 6° for the upper surface and -5° for the loser surface. Their results of this study compared well with those of other experiments and computational results.
As the potential of civil and military helicopters has increased, more complex and demanding missions in increasingly hostile environments have been required. Users, designers, and manufacturers have an urgent need for information about human behavior and function to create systems that take advantage of human capabilities, without overloading them. Because there is a large gap between what is known about human behavior and the information needed to predict pilot workload and performance in the complex missions projected for pilots of advanced helicopters, Army and NASA scientists are actively engaged in Human Factors Research at Ames. The research ranges from laboratory experiments to computational modeling, simulation evaluation, and inflight testing. Information obtained in highly controlled but simpler environments generates predictions which can be tested in more realistic situations. These results are used, in turn, to refine theoretical models, provide the focus for subsequent research, and ensure operational relevance, while maintaining predictive advantages. The advantages and disadvantages of each type of research are described along with examples of experimental results.
A method is developed to predict the rotor hover performance. This method solves the compressible mass conservation equation much like current full potential codes and can therefore predict the transonic flows on a rotor. However, the newly developed approach also allows for the free convection of shed vorticity and permits the computation of the entire hover wake system. The method uses a vortex embedding scheme in potential flow and has been implemented in a computer code, HELIX -I. To predict power we implement a simple boundary layer and two different induced-drag integration schemes. The induced-drag is obtained from surface pressure integration and an energy flux integral. Comparisons between computations and experiment show good agreement for the prediction of power polars, surface pressure distribution, and tip vortex geometry.
Three models have been applied to very rapid compression of oxygen in a dead-ended tube. Pressures as high as 41 MPa (6000 psi) leading to peak temperatures of 1400 K are predicted. These temperatures are well in excess of the autoignition temperature (750 K) of teflon, a frequently used material for lining hoses employed in oxygen service. These findings are in accord with experiments that have resulted in ignition and combustion of the teflon, leading to the combustion of the stainless steel braiding and catastrophic failure. The system analyzed was representative of a capped off-high-pressure oxygen line, which could be part of a larger system. Pressurization of the larger system would lead to compression in the dead-end line, and possible ignition of the teflon liner. The model consists of a large plenum containing oxygen at the desired pressure (500 to 6000 psi). The plenum is connected via a fast acting valve to a stainless steel tube 2 cm inside diameter. Opening times are on the order of 15 ms. Downstream of the valve is an orifice sized to increase filling times to around 100 ms. The total length from the valve to the dead-end is 150 cm. The distance from the valve to the orifice is 95 cm. The models describe the fluid mechanics and thermodynamics of the flow, and do not include any combustion phenomena. A purely thermodynamic model assumes filling to be complete upstream of the orifice before any gas passes through the orifice. This simplification is reasonable based on experiment and computer modeling. Results show that peak temperatures as high as 4800 K can result from recompression of the gas after expanding through the orifice. An approximate transient model without an orifice was developed assuming an isentropic compression process. An analytical solution was obtained. Results indicated that fill times can be considerably shorter than valve opening times. The third model was a finite difference, 1-D transient compressible flow model. Results from the code show the recompression effect but predict much lower peak temperatures than the thermodynamic model.
This paper describes and compares three parallel algorithms for solving sparse triangular systems of equations. These methods involve some preprocessing overhead and are primarily of interest in solving many systems with the same coefficient matrix. The first approach is to use a fixed blocksize and form the inverse of the diagonal blocks. The second approach is to use a variable blocksize and reorder the unknowns so that the diagonal blocks are diagonal matrices. The latter technique is called level scheduling because of how it is represented in the adjacency graph, and both row-wise and jagged diagonal storage for the off-diagonal blocks are considered. These techniques are analyzed for general parallel computers and experiments are presented for the eight-processor Alliant FX/8.
The three-dimensional (3-D) instability of a two-dimensional (2-D) flow with elliptical streamlines has been proposed as a generic mechanism for the breakdown of many 2-D flows. A physical interpretation for the mechanism is presented together with an analytical treatment of the problem. It is shown that the stability of an elliptical flow is governed by an Ince equation. An analytical representation for a localized solution is given and establishes a direct link with previous computations and experiments.
The fundamental fluid dynamics governing instability and transition to turbulence in boundary layers are considered, and attention is focused on the key aspects of nonlinear dynamics central to the transition process and to turbulent boundary-layer phenomena. Emphasis is placed on truly nonlinear theories, in which the boundary layer mean-flow profile is completely altered from its original form. Nonlinear TS transitions, Euler-stage interactions, and vortex/wave interactions are discussed, and compressible boundary layers are analyzed. Connections with experiments and computations are outlined, along with overall trends including the extension of the nonlinear theory and the advancement in the compressible and other flow regimes.
The aerodynamic cofficients and trim angle for an aerobrake at Mach 9.2 and 11.8 were found using a combination of experiment and computation. Free-flight tests were performed at NASA Ames Research Center's Hypervelocity Free-Flight Aerodynamic Facility, and the forebody pressure distribution was calculated using a three-dimensional Navier-Stokes code with an effective specific heat ratio. Using the computed drag, lift, and moments to prescribe the number of terms in the aerodynamic coefficient expansions and to specify the values of the higher order terms, the experimental aerodynamic coefficients and trim angle were found using a six-degree-of-freedom, weighted, least-squares analysis. The experimental and computed aerodynamic coefficients and trim angles are in good agreement. The trim angle obtained from the free-flight tests, 14.7 deg, differs from the design trim angle, 17 deg, and from the Langley wind tunnel results, 12 deg in air and 17 deg in CF4. These differences are attributable to real-gas effects.
Experimental interferograms, schlieren, and shadow-graphs are routinely used for validating ideal- and real-gas computational fluid dynamics (CFD) codes. Interferograms are used for quantitative comparisons of integrated density. Schlieren and shadowgraphs are used for comparing the locations of flow field features; the types of flowfield features recorded in these images include shocks, shear layers, and expansion fans. By constructing images from flow field simulations, one-to-one comparisons of computation and experiment are possible. During the period from 1 May 1991 to 31 Jan. 1992, the emphasis of this research was on developing techniques for constructing these images from flow field simulations.
The purpose of this study is to characterize the scattering effects of a solar panel on the Space Station Space-to-Ground Subsystem reflector antenna when the antenna is operated in the RF tracking (difference) mode. This analysis is based on a combination of geometrical theory of diffraction and aperture integration techniques. The advantage of this combination method is its capability in predicting reflector antenna patterns not only in the forward region but also in the far-out sidelobes and backlobes. To verify the analytical model, measurements were performed on the Johnson Space Center far-field antenna test facility. Good agreement between computed and measured results was obtained. It was shown from computation and experiment that the solar panel scattering interference causes a shift in null position, a decrease in the depth of the null, as well as a decrease in the gain on the antenna difference mode patterns.
The role of experiment in Computational Fluid Dynamics (CFD) is discussed. Flow modeling of complex physics and determination of accuracy limits (confidence) are two ways in which experimentation can be used to develop CFD. The results of this discussion are presented in viewgraph form.
Numerical flowfields around a scramjet inlet model are simulated and analyzed. The present inlet flowfield is characterized by thick boundary-layer ingestion and strong viscous/inviscid interaction because of a combined effect of high hypersonic freestream Mach and low Reynolds numbers. Shock-induced separation further enlarges regions of viscous flows which occupy most of the inlet flowfield. Results obtained from the computations with the PARC code developed for ideal gas are presented for several 2D cases at various hypersonic Mach numbers ranging from 10 to 25, and two 3D simulations at Mach numbers of 12 and 19 are also discussed. Comparison between computation and experiment is made in terms of pressure distributions at the wall center line. Large discrepancy is observed and may be partially attributed to the lack of real gas and/or 3D effects in the simulation as well as to the uncertainty of the experiment.
Optical images, such as experimental interferograms, schlieren, and shadowgraphs, are routinely used to identify and locate features in experimental flow fields and for validating computational fluid dynamics (CFD) codes. Interferograms can also be used for comparing experimental and computed integrated densities. By constructing these optical images from flow-field simulations, one-to-one comparisons of computation and experiment are possible. During the period from February 1, 1992, to November 30, 1992, work has continued on the development of CISS (Constructed Interferograms, Schlieren, and Shadowgraphs), a code that constructs images from ideal- and real-gas flow-field simulations. In addition, research connected with the automated film-reading system and the proposed reactivation of the radiation facility has continued.
Stanford Telecom developed the Three-Dimensional Event-Driven Graphics Environment (3D-EDGE) for NASA GSFC's (GSFC) Communications Link Analysis and Simulation System (CLASS). 3D-EDGE consists of a library of object-oriented subroutines which allow engineers with little or no computer graphics experience to programmatically manipulate, render, animate, and access complex three-dimensional objects.