Engineering Papers⌕ Search

Engineering topics

Bird, G. A.

Publications and source records attributed to Bird, G. A..

At least 19 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS1992

DSMC analysis in a heterogeneous parallel computing environment

A methodology for implementing parallel DSMC codes in a heterogeneous computing environment is described. The methodology involves the use of a common message-passing software library together with recently developed software that handles the actual interprocessor communications in a standard manner across a variety of computing platforms. Benchmark tests using a simple DSMC model problem were performed on an Intel iPSC/860, a Cray-YMP and a group of Sun workstations. The approach was found to give speedups that scaled linearly with problem size on all the computing platforms tested. This methodology was then incorporated into a production-type DSMC code to allow the simulation of problems that would not otherwise have been practical. The application of this production code to simulations of hypersonic shear flows and shock-lip interactions under near-continuum conditions is described. Synchronous and asynchronous models for implementing parallelism into DSMC simulations are also described and both models are shown to produce the same steady-state result.

Wilmoth, R. G.↗
NASA NTRS1991

Shock wave-boundary layer interactions in rarefied gas flows

A numerical study is presented, using the direct simulation Monte Carlo (DSMC) method, of shock wave-boundary layer interactions in low density supersonic flows. Test cases include two-dimensional, axially-symmetric and three-dimensional flows. The effective displacement angle of the boundary layer is calculated for representative flat plate, wedge, and cone flows. The maximum pressure, shear stress, and heat transfer in the shock formation region is determined in each case. The two-dimensional reflection of an oblique shock wave from a flat plate is studied, as is the three-dimensional interaction of such a wave with a sidewall boundary layer.

Bird, G. A.↗
NASA NTRS1991

An efficient DSMC algorithm applied to a delta wing

A new algorithm for 3D direct simulation Monte Carlo (DSMC) is tested and numerical results are compared with wind tunnel data and results obtained earlier with a more traditional DSMC code. The test case is the flowfield around a delta wing at incidence at Knudsen number of 0.016 and Mach number of 20.2. The results are shown to compare favorably with both experimental and earlier numerical results. The new algorithm is described with special emphasis placed on its distinctive features: Cartesian/unstructured combination grid, special body surface definition, discretization in physical space.

Rault, D. F. G.↗
NASA NTRS1991

Efficiency and discrepancy in the direct simulation methods

The paper reviews the factors that affect the versatility, computational efficiency, and accuracy of the various direct simulation methods that are being applied to problems in rarefied gas dynamics. The influence of computer hardware developments on the cost-effectiveness is discussed, as are the software factors that influence the speed with which simulation programs can be applied to new problems. Particular attention is given to a novel method that improves the efficiency of the simulation for complex 3D problems. The topical issues of reduced resolution procedures, stored lists, and variance reduction schemes are discussed in some detail.

Bird, G. A.↗
NASA NTRS1990

Efficiency and discrepancy in the direct simulation methods

The paper reviews the factors that affect the versatility, computational efficiency, and accuracy of the various direct simulation methods that are being applied to problems in rarefied gas dynamics. The quality of the simulation can be affected by the underlying motivation and objectives. The influence of computer hardware developments on the cost-effectiveness is discussed, as are the software factors that influence the speed with which simulation programs can be applied to new problems. Particular attention is given to a novel method that improves the efficiency of the simulation for complex three-dimensional problems. The topical issues of reduced resolution procedures, stored lists, and variance reduction schemes are discussed in some detail.

Bird, G. A.↗
NASA NTRS1990

Application of the direct simulation Monte Carlo method to the full shuttle geometry

A new set of programs has been developed for the application of the direct simulation Monte Carlo (or DSMC) method to rarefied gas flows with complex three-dimensional boundaries. The programs are efficient in terms of the computational load and also in terms of the effort required to set up particular cases. This efficiency is illustrated through computations of the flow about the Shuttle Orbiter. The general flow features are illustrated for altitudes from 170 to 100 km. Also, the computed lift-drag ratio during re-entry is compared with flight measurements.

Bird, G. A.↗
NASA NTRS1990

Influence of local configuration on the backflow from small rocket thrusters

A numerical study has been made of the low density plume flow from a small thruster motor. For the case studied, the addition of a thin plume shield with a diameter 1.67 times the nozzle exit diameter was found to lead to a reduction in the backflow by several orders of magnitude. Calculations were also made for the three-dimensional flow produced by a pair of motors, in order to determine the increase in the backflow due to the interaction of the plumes. The calculations were made with the direct simulation Monte Carlo (DSMC) method, and one of the purposes of the work was to test a new and faster set of algorithms for this method.

Bird, G. A.↗
NASA NTRS1989

Computation of electron density in high altitude re-entry flows

The direct simulation Monte Carlo (DSMC) method has been modified to deal with the very small degree of ionization that occurs in the early stages of the 'communications blackout' phase of re-entry. The simplified model of the process assumes that the ionization is entirely a consequence of the reaction N + O yields NO(+) + e-. The measurements from the RAM C-II test vehicle at an altitude of 81 km have been chosen as the test case. It was found that the form of the electron distribution along the shock layer was in agreement with the measurements. That normal to the surface was also in agreement with the measurements if the surface is regarded as fully catalytic. However, quantitative agreement for the fully catalytic case requires a reaction rate about an order of magnitude greater than those generally associated with this reaction.

Bird, G. A.↗
NASA NTRS1989

Grids for DSMC computation of two and three-dimensional flows

Contemporary grid schemes for use in Direct Simulation Monte Carlo (DSMC) method of computing two- and three-dimensional flows are briefly surveyed. The grids described include: (1) simple, constant area cells defined by a rectangular grid; (2) multilevel rectangular grids; (3) adaptive rectangular cells; (4) 'point reference' scheme; (5) body-fitted 'analytical' cells; (6) 'general' schemes; (7) 'general' schemes with subcells; (8) orthogonal transformation to rectangular grid; (9) structural tetrahedral grids; (10) rectangular multilevel subcells with adaptive body-fitted cells.

Bird, G. A.↗
NASA NTRS1989

Perception of numerical methods in rarefied gasdynamics

The relationships between various numerical methods applied to problems in rarefied gasdynamics are discussed, with emphasis on conflicting viewpoints and computational requirements associated with physical simulation versus the numerical solution of the Boltzmann equation. The basic differences between the molecular dynamics and direct simulation methods are shown to affect their applicability to dense and rarefied flows. Methods for the probabilistic selection of representative collision in the direct simulation Monte Carlo method are reviewed. A method combining the most desirable features of the earlier methods is presented.

Bird, G. A.↗
NASA NTRS1988

Thermal and pressure diffusion effects in high altitude flows

An apparently anomalous change in composition adjacent to a non-catalytic surface in an earlier DSMC calculation of a non-reacting gas mixture is shown to have been due to thermal diffusion. Both thermal and pressure diffusion can lead to species separation effects in the flow of gas mixtures, but are not generally included in Navier-Stokes formulations. Additional DSMC calculations are made for blunt-body and plume expansion flows using both a simple gas and an otherwise identical gas mixture. These test calculations indicate that the pressure distribution on the forward face of a blunt body is largely unaffected by the species separation in the stagnation region. The gas density in this region is significantly affected and there is a slight change in the heat transfer to the surface. The pressure distribution around the shoulder of the body is affected. Very large separations occur in the plume expansion of the gas mixtures and there is an order of magnitude increase in density in the backflow region for the mixture.

Bird, G. A.↗
NASA NTRS1987

Nonequilibrium radiation during re-entry at 10 km/s

The direct simulation Monte Carlo method, including a real air model with thermal radiation, is applied to the flows associated with the two sets of measurements that are directly relevant to the projected aeroassisted orbital transfer vehicle. The first is a shock tube measurement of the radiation from a 10 km/s shock wave in air that was made at AVCO in 1962. The second is the flight data that was obtained from the Project Fire re-entry test vehicles in 1964. The calculations for both cases were made with a program that models the one-dimensional flow along a stagnation streamline. The shock standoff distance for the Fire vehicle was obtained from the theoretical studies that were associated with its launch. The simulation employed a partly phenomenological model for the nonequilibrium radiation. It was found that the results from the calculation were consistent with the measured radiation in each case, and also with the convective heat transfer data for the Fire vehicle. The uncertainties associated with the spectral absorptance and recombination probability at the surface appear to be as serious as those associated with the reaction rates.

Bird, G. A.↗
NASA NTRS1987

Direct simulation of high-vorticity gas flows

The computational limitations associated with the molecular dynamics (MD) method and the direct simulation Monte Carlo (DSMC) method are reviewed in the context of the computation of dilute gas flows with high vorticity. It is concluded that the MD method is generally limited to the dense gas case in which the molecular diameter is one-tenth or more of the mean free path. It is shown that the cell size in DSMC calculations should be small in comparison with the mean free path, and that this may be facilitated by a new subcell procedure for the selection of collision partners.

Bird, G. A.↗
NASA NTRS1986

Direct simulation of typical AOTV entry flows

The direct simulation Monte Carlo method is applied to two flows that are relevant to the projected aeroassisted orbital transfer vehicle. The first is the axially symmetric flow about a 70 degree blunted cone shield with a cylindrical afterbody, and the second is the one-dimensional flow along the stagnation streamline of the first flow. The shock standoff distance for the stagnation streamline study is obtained from the full axially symmetric calculation. Advantage is then taken of the speed of the one-dimensional calculation to study the sensitivity of the stagnation point heat transfer to a number of physical variables. These include the chemical reaction rate coefficients, the surface catalycity, the electric field associated with ambipolar diffusion, and the nonequilibrium thermal radiation from the shock front.

Bird, G. A.↗
NASA NTRS1986

The diffusion of individual molecules within a gas

The Direct Simulation Monte Carlo method is used to study the positional history of the individual molecules in a gas that is homogeneous at the macroscopic level and is in Maxwellian equilibrium at the microscopic level. The behavior at small times is characterized by 'persistence of velocity' effects, and a 'random walk' type of dispersal occurs over a longer timescale. It is shown that the rate of dispersal can be directly related to the self-diffusion coefficient. In addition, the diffusion coefficients are obtained directly from one-dimensional calculations, and the local Knudsen number at which the Chapman-Enskog theory breaks down is determined. Results are presented for both simple gases and gas mixtures.

Bird, G. A.↗
NASA NTRS1985

Rocket nozzle lip flow by direct simulation Monte Carlo method

The flow in the immediate vicinity of a rocket nozzle lip has been analyzed, and the results are presented. A rapid change in gas composition is observed in the flow around the lip depending principally on species molecular mass and local flow angle. The divergence of axial and radial temperatures indicating breakdown of translational equilibrium is observed in the results.

Hueser, J. E.↗
NASA NTRS1985

Monte Carlo simulations in support of the Shuttle upper atmospheric mass spectrometer experiment

This paper presents the results of calculations obtained with a Direct Simulation Monte Carlo (DSMC) method that describes both the external flow about the nose region of the Shuttle Orbiter and the internal flow in an inlet tube that provides the initial path between the shock-processed gases and a mass spectrometer mounted inside the Orbiter. A dedicated, three-dimensional version of the DSMC was developed during this study for the internal flow simulation along with the procedure for interfacing the external and internal flows. The calculations span the 140 to 95 km altitude range, which includes most of the hypersonic transitional flow encountered during reentry. All calculations were for a multicomponent gas mixture consisting of five chemical species while simulating the effects of transitional, rotational, vibrational, and chemical nonequilibrium. The results of the calculations show that within the entry region of the inlet tube where the gas has equilibrated with the sidewall tube temperature, the pressure is substantially less than the pressure at the external surface. This pressure correction for the entry region is significant for all conditions investigated and increases with altitude. The results highlight the structure of both the external and internal flow. Results of parametric studies show the sensitivity of the equilibrated inlet sidewall pressure to mass flow rates, gas-surface reflection model, tube side-wall temperature variations, and surface recombinations.

Moss, J. N.↗