SEARCH · Engineering Papers
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.
The effectiveness of fluid containment near an interior stagnation point and within a self-confined stagnation region is determined by numerically solving the species conservation equation for a bi-component mixture. The flow geometry is that of a swirling fluid stream containing a stationary eddy on the axis of rotation. The base flow is axisymmetric, and the Reynolds number is equal to 50. Schmidt numbers range from 0.1 to 10.
The reported investigation is concerned with mesoscale squall-line simulation utilizing numerical techniques. Particular attention is given to fourth-order results and their implications for the dynamical evolution of mesoscale squall-line systems which contain severe local storms. An adiabatic inviscid set of prognostic equations is utilized in a z coordinate system for the fundamental experiment. The fourth-order advection is employed for all time-dependent equations. The Euler-backward time marching scheme is utilized with a 60 second time step. The horizontal mesh length is 42 km. Horizontal diffusion is accomplished by utilizing a smoother-desmoother. Lateral boundary conditions are designed to reduce the development of strong gradients in dependent variables near the boundaries by bringing interior values from the grid to the boundaries.
This paper presents several improvements to the BDP model of steady-state burning of AP composite solid propellants. The Price-Boggs-Derr model of AP monopropellant burning is incorporated to represent the AP. A separate energy equation is written for the binder to permit a different surface temperature from the AP; this includes an analysis of the sharing of primary diffusion flame energy, and correction of a BDP model inconsistency in treating the binder regression rate. A method for assembling component contributions to calculate the burning rates of multimodal propellants is also presented. Results are shown in the form of representative burning rate curves, comparisons with data, and calculated internal details of interest. Ideas for future work are discussed in an Appendix.
Different distributions of variables over the horizontal array of grid points in an ocean circulation model are investigated, using the shallow water equations as a guide in the choice of finite-difference schemes for use in ocean modeling. It is shown that the scheme with diffusive dissipation, in which the horizontal velocity is carried at the center and the height field is carried at each corner of a rectangular grid, successively suppresses numerical noise in a coarse (greater than 100 km) grid ocean model. For resolutions smaller than 50 km, it is shown that the scheme in which zonal velocity is carried at points to the east and west of the point of a rectangular grid where the height is carried, with meridional velocity carried to the north and south of the height point, can be free of noise for the gravest mode.
Accurate numerical simulations of turbulent free shear flows were performed in order to study the behavior of certain critical turbulence quantities which have been difficult to measure in the laboratory. The three-dimensional time-dependent nonlinear Navier-Stokes equations are solved by the direct numerical simulation technique without turbulence modeling or ad hoc closure assumptions. The pressure interaction terms were factored into diffusion and return-to-isotropy components according to two models. Simulations of the evolution of a turbulent wake with and without swirl present were performed at sufficiently high Reynolds numbers to insure that the flows were strongly nonlinear and turbulent, but low enough so that all scales of motion containing significant energy are adequately resolved. It is noted that the presence of swirl tends to enhance the amplitude of the rms pressure levels.
A new analytical model for chemoviscosity variation during cure of thermosetting resins was developed. This model is derived by modifying the widely used WLF (Williams-Landel-Ferry) Theory in polymer rheology. Major assumptions involved are that the rate of reaction is diffusion controlled and is linearly inversely proportional to the viscosity of the medium over the entire cure cycle. The resultant first order nonlinear differential equation is solved numerically, and the model predictions compare favorably with experimental data of EPON 828/Agent U obtained on a Rheometrics System 4 Rheometer. The model describes chemoviscosity up to a range of six orders of magnitude under isothermal curing conditions. The extremely non-linear chemoviscosity profile for a dynamic heating cure cycle is predicted as well. The model is also shown to predict changes of glass transition temperature for the thermosetting resin during cure. The physical significance of this prediction is unclear at the present time, however, and further research is required. From the chemoviscosity simulation point of view, the technique of establishing an analytical model as described here is easily applied to any thermosetting resin. The model thus obtained is used in real-time process controls for fabricating composite materials.
Numerical techniques to calculate the thermal and geometric properties of line emission from H II 'blister' regions are presented. It is assumed that the density distributions of the H II regions are a function of two dimensions, with rotational symmetry specifying the shape in three-dimensions. The thermal and ionization equilibrium equations of the problem are solved by spherical modeling, and a spherical sector approximation is used to simplify the three-dimensional treatment of diffuse ionizing radiation. The global properties of H II 'blister' regions near the edges of a molecular cloud are simulated by means of the geometry/density distribution, and the results are compared with observational data. It is shown that there is a monotonic increase of peak surface brightness from the i = 0 deg (pole-on) observational position to the i = 90 deg (edge-on) position. The enhancement of the line peak intensity from the edge-on to the pole-on positions is found to depend on the density, stratification, ionization, and electron temperature weighting. It is found that as i increases, the position of peak line brightness of the lower excitation species is displaced to the high-density side of the high excitation species.
A time-dependent, three-dimensional, multi-ion model of the ionospheric F region at 120-800 km altitude is presented. Account is taken of field-aligned diffusion, cross-field electrodynamic drifts in equatorial and high latitude regions, interhemispheric flow, thermospheric winds, polar wind escape, energy-dependent chemical reactions and neutral composition changes. Attention is also given to the effects of ion production by solar EUV radiation and auroral precipitation, thermal conduction, diffusion-thermal heat flow, local heating and cooling processes, offsets between the geomagnetic and geographic poles, and bending of field lines near the magnetic equator. The model incorporates all phenomena described by previous models and can be applied to tracing magnetic storm and substorm disturbances from high to low latitudes on a global scale. Sample results are provided for ionospheric features during a June solstice, the solar maximum and in a period of low geomagnetic activity. The model will eventually be used to study coupled ionosphere-thermosphere activity.
A numerical investigation of two-dimensional unsteady boundary layer in a transonic diffuser flow with self-excited oscillations and strong flow separation by solving the compressible, Reynolds-averaged, thin-layer Navier-Stokes equations with two-equations turbulence model is described. Three different meshes with constant streamwise mesh distribution and varying vertical mesh distribution were used. Results obtained indicate that a refinement of mesh studied here has minimal effect on the mean boundary layer flow but significantly increases the amplitude of oscillation of all flow variables. Comparisons of unsteady wall pressure, velocity profile, terminal shock, and separation pocket among computations and with experiment are presented.
Data related to the development of cosmic rays are discussed. The relationship between cosmic ray production and the steady-state Boltzmann equation is analyzed. The importance of the power-law spectrum, the scattering rate, the theory of shock acceleration, anisotropic instabilities, and cosmic ray diffusion in the formation of cosmic rays is described. It is noted that spacecraft observations at the earth's bow shock are useful for studying cosmic rays and that the data support the collisionless shock-wave theory of cosmic ray origin.
The temporal evolution of Langmuir waves excited by high-power, high-frequency (HF) radio waves in the ionosphere is studied theoretically. This study is motivated by past observations made with the 450 MHz radar at Arecibo Observatory in Puerto Rico. Two kinds of nonlinear damping to the parametric decay instability are considered in the derivation of the rate equation for the spectral intensity of enhanced Langmuir waves. They are Langmuir wave cascading caused by nonlinear Landau damping and cross-field electron diffusion. The first damping process leads to the saturation of individual unstable Langmuir wave. The second process, which results from the incoherent scattering of electron orbits by the total excited Langmuir waves, yields anomalous damping that applies to each Langmuir wave. Consequently, Langmuir waves with smaller growth rates will be suppressed by those with larger growth rates. Such a mode competition process may cause the overshoot of the HF-enhanced plasma line observed with the Arecibo 430 MHz radar. Favorable agreement is obtained between theory and the Arecibo observations.
An analysis is given for the entrainment of dust into a turbulent radial wall jet. Equations are solved based on incompressible flow of a radial wall jet into which dust is entrained from the wall and transported by turbulent diffusion and convection throughout the flow. It is shown that the resulting concentration of dust particles in the flow depends on the difference between the applied shear stress at the surface and the maximum level of shear stress that the surface can withstand (varies as rho(sub d)a(sub g)D) i.e., the pressure due to the weight of a single layer of dust. The analysis is expected to have application to the downflow that results from helicopter and VTOL aircraft.
A review is presented of methods in which composite or reduced Navier-Stokes (RNS) equations are treated with a pressure-gradient-based flux-vector splitting. The methods are similar to large Re asymptotic formulations, and streamwise diffusion terms are ignored in favor of an explicit deferred corrector based on higher-order diffusion terms. The methods can be used for 2D and 3D supersonic flows in which the effects of real gas are incorporated. Several examples of the procedure are given, and subsonic and supersonic flows are handled well with relaxation procedures that incorporate multigrid acceleration. Effective solutions are described for problems ranging from incompressible flows and supersonic flows to sharp shocks and reverse-flow capturing.
X-ray photoelectron spectroscopy was employed, in conjunction with ion bombardment, to analyze the chemical composition profile across thin (less than 50 nm) oxide films on chemically vapor deposited Si3N4. The thermal oxides were grown in dry oxygen at 1100 C on samples with or without native oxide film (formed in room air). The results show that the thermal oxidation product was silicon oxynitride of graded N:O ratio, and that the presence of a native oxide film promotes the formation of a SiO2 crust over the oxynitride. It is proposed that the fundamental mechanism of Si3N4 oxidation is progressive O-for-N substitution in the silicon oxynitride unit tetrahedron, which is best designated SiN(2-x)O(2+x), where x is also an index of depth. The corresponding equation for nonstoichiometric oxidation of Si3N4 describes a bulk (rather than an interface) reaction process, with significant implications for O2 and N2 fluxes and diffusivities.
Spatial and seasonal distribution of ozone and other trace gases are simulated using a interactive two-dimensional model of the stratosphere updated to include full gas phase chemistry. The model consists of a primitive equation dynamics module, a full radiative transfer scheme, and a comprehensive gas phase chemistry module. The circulation is derived from heating rates in the stratosphere that are calculated using model-generated ozone. In the troposphere, parameterized heating rates are adopted. The eddy momentum flux divergence in the zonal mean momentum equation is given by the eddy fluxes of potential vorticity. Eddy fluxes of potential vorticity and tracers are parameterized using a set of predetermined diffusion coefficients. The adopted values for K(sub yy), show a hemispherical asymmetry in that the values in the lower stratosphere are consistently smaller in the southern hemisphere. The asymmetry in K(sub yy) and in the parameterization of the tropospheric heating rate results in an asymmetry in the circulation giving rise to unique signatures in the trace gas distributions. The model successfully simulates the observed asymmetry in the column abundance of the springtime ozone maxima between northern and southern hemisphere. Results for other trace gases are in agreement with the gross observed features although specific differences exist.
Fluxes of electrons and protons in Jupiter's radiation belts are calculated with the source (radial diffusion inward from the solar wind) and the loss (synchroton radiation). The calculations are tested against the measured radio-wave wavelength distribution, the radio-wave distribution with distance from Jupiter, and the degree of polarization of the radio waves. The Fokker-Planck equation is solved by using the method of Farley and Walt (1971) with the fixed flux at the outer boundary and the zero flux at Jupiter's surface. It is found, in agreement with Brice and McDonough (1973), Jacques and Davis (1973), and Birmingham et al. (1974), that the usual magnetic and electric diffusions, which vary as L to the 10th power and L to the 6th power, respectively, are insufficient to furnish the required electrons and that the diffusion driven by ionospheric winds of Brice and McDonough is strong enough to furnish Jupiter's belts. An additional loss mechanism close to Jupiter is required to remove the electrons, particularly those at low energies.
A numerical investigation is reported of turbulent incompressible jets confined in two ducts, one cylindrical and the other conical with a 5 degree divergence. In each case, three Craya-Curtet numbers are considered which correspond, respectively, to flow situations with no moderate and strong recirculation. Turbulence closure is achieved by using the k-epsilon model and a recently proposed realizable Reynolds stress algebraic equation model that relates the Reynolds stresses explicitly to the quadratic terms of the mean velocity gradients and ensures the positiveness of each component of the turbulent kinetic energy. Calculations are carried out with a finite-volume procedure using boundary-fitted curvilinear coordinates. A second-order accurate, bounded convection scheme and sufficiently fine grids are used to prevent the solutions from being contaminated by numerical diffusion. The calculated results are compared extensively with the available experimental data. It is shown that the numerical methods presented are capable of capturing the essential flow features observed in the experiments and that the realizable Reynolds stress algebraic equation model performs much better than the k-epsilon model for this class of flows of great practical importance.
A physical model for a thruster discharge was developed, consisting of a spatially diverging plasma sustained electrically between a small ring cathode and a larger ring anode in a cylindrical chamber with an axial magnetic field. The associated boundary-value problem for the coupled partial differential equations with mixed boundary conditions, which describe the electric potential and the plasma velocity fields, was solved in closed form. By means of quantum-mechanical perturbation theory, a formula for the number S(E) of atoms sputtered on the average by an ion of energy E was derived from first principles. The boundary-value problem describing the diffusion of the sputtered atoms through the surrounding rarefied electron-ion plasma to the system surfaces of ion propulsion systems was formulated and treated analytically. It is shown that outer boundary-value problems of this type lead to a complex integral equation, which requires numerical resolution.