Engineering topics

Spalart, P. R.

Publications and source records attributed to Spalart, P. R..

35 records · Page 2

NASA NTRS1987

Self similarity of two point correlations in wall bounded turbulent flows

The structure of turbulence at a height y from a wall is affected by the local mean shear at y, by the direct effect of the wall on the eddies, and by the action of other eddies close to or far from the wall. Some researchers believe that a single one of these mechanisms is dominant, while others believe that these effects have to be considered together. It is important to understand the relative importance of these effects in order to develop closure models, for example for the dissipation or for the Reynolds stress equation, and to understand the eddy structure of cross correlation functions and other measures. The specific objective was to examine the two point correlation, R sub vv, of the normal velocity component v near the wall in a turbulent channel flow and in a turbulent boundary layer. The preliminary results show that even in the inhomogeneous turbulent boundary layer, the two-point correlation function may have self similar forms. The results also show that the effects of shear and of blocking are equally important in the form of correlation functions for spacing normal to the wall. But for spanwise spacing, it was found that the eddy structure is quire different in these near flows. So any theory for turbulent structure must take both these effects into account.

Hunt, J. C. R.↗

NASA NTRS1987

A general form for the dissipation length scale in turbulent shear flows

It has been found that, for a wide range of turbulent wall-bounded shear flows with mean velocity profile U(y), the scale length L-epsilon determining the dissipation is approximately described in terms of distance from the wall (y), the mean shear (dU/dy), and the variance of the normal component of turbulence (bar u-squared), by the formula: L-epsilon to the -1 power is approximately equal to (A sub B / y) + A sub S (dU/dy / square root of bar u-squared) where L-epsilon = epsilon/(bar u-squared) to the 3/2 power. To match with shear-free boundary layers, A sub B is approximately 0.27, and to match with the log layer, A sub S is approximately 0.46. The shear flows tested here were: boundary layers over a flat plate, sink flow, oscillatory flow, and channel flow. The use of the square root of bar u-squared as a velocity scale minimizes the effects of Reynolds number. However, the formula fails within a distance of order L-epsilon for the regions where dU/dy = 0.

Hunt, J. C. R.↗

NASA NTRS1987

Active layer model for wall-bounded turbulence

The active-layer model for wall-bounded turbulence hypothesizes that the nonlinear terms are large only in a thin layer near the wall, and hence the turbulence in the region outside the active inner layer can be modeled as a linear fluctuating flow driven by the active layer. This hypothesis is tested using data obtained from a direct simulation of turbulent channel flow. It is found that the nonlinear effects are the strongest near the wall with a maximum at around y sup + = 20 and, outside the near-wall region, these involve primarily the cascading mechanism leading to dissipation.

Landahl, M. T.↗

NASA NTRS1987

A full simulation of a vortex ring

A three-dimensional spectral method is developed for the solution to the incompressible Navier-Stokes equations in an unbounded domain. The spectral method relies on divergence-free basis functions as proposed by Leonard (1981). The basis functions are formed using vector spherical harmonics and Jacobi polynomials together with a mapping in the radial direction. An axisymmetric code was written and is verified using an exact solution of the Stokes equations. Preliminary results for the evolution of a vortex ring according to the Navier-Stokes equations are presented.

Stanaway, S. K.↗

NASA NTRS1986

Internal structure of a vortex breakdown

An axisymmetric vortex breakdown was well simulated by the vortex filament method. The agreement with the experiment was qualitatively good. In particular, the structure in the interior of the vortex breakdown was ensured to a great degree by the present simulation. The second breakdown, or spiral type, which occurs downstream of the first axisymmetric breakdown, was simulated more similarly to the experiment than before. It shows a kink of the vortex filaments and strong three-dimensionality. Furthermore, a relatively low velocity region was observed near the second breakdown. It was also found that it takes some time for this physical phenomenon to attain its final stage. The comparison with the experiment is getting better as time goes on. In this paper, emphasis is placed on the comparison of the simulated results with the experiment. The present results help to make clear the mechanism of a vortex breakdown.

Nakamura, Y.↗

NASA NTRS1986

Numerical simulation of boundary layers. Part 1: Weak formulation and numerical method

A numerical method designed to solve the time-dependent, three-dimensional, incompressible Navier-Stokes equations in boundary layers is presented. The fluid domain is the half-space over a flat plate, and periodic conditions are applied in the horizontal directions. The discretization is spectral. The basis functions are divergence-free and a weak formulation of the momentum equation is used, which eliminates the pressure term. An exponential mapping and Jacobi polynomials are used in the semi-infinite direction, with the irrotational component receiving special treatment. Issues related to the accuracy, stability and efficiency of the method are discussed. Very fast convergence is demonstrated on some model problems with smooth solutions. The method has also been shown to accurately resolve the fine scales of transitional and turbulent boundary layers.

Spalart, P. R.↗

NASA NTRS1985

Vortex breakdown simulation

A vortex breakdown was simulated by the vortex filament method, and detailed figures are presented based on the results. Deformations of the vortex filaments showed clear and large swelling at a particular axial station which implied the presence of a recirculation bubble at that station. The tendency for two breakdowns to occur experimentally was confirmed by the simulation, and the jet flow inside the bubble was well simulated. The particle paths spiralled with expansion, and the streamlines took spiral forms at the breakdown with expansion.

Nakamura, Y.↗

NASA NTRS1985

Simulation of rotating stall by the vortex method

The vortex method, coupled to a boundary-layer solver, is applied to the numerical simulation of high Reynolds number incompressible flow in two-dimensional cascades. Periodic conditions are imposed along the plane of the cascade, with several blades per period. Good agreement is found with two finite-difference methods for a single-blade case. When a staggered cascade is treated with five independent blades, the simulation predicts rotating stall, for a range of angles of attack and stagger, and the essential features of the flow are correct. The stall cell steadily propagates along the cascade. The sensitivity of this phenomenon to two parameters is studied, and the stall boundary is found. Quantitative results and visualizations are presented.

Spalart, P. R.↗

NASA NTRS1985

Direct numerical simulation of equilibrium turbulent boundary layers

This paper describes the simulation of turbulent boundary layers by direct numerical solution of the three-dimensional, time-dependent Navier-Stokes equations, using a spectral method. The flow is incompressible, with Re sub delta = 1000. The equations are written in the self-similar coordinate system and periodic streamwise and spanwise boundary conditions are imposed. A family of nine 'equilibrium' boundary layers, from the strongly accelerated 'sink' flow to Stratford's separating boundary layer is treated. Good general agreement with experiments is observed. The effects of pressure gradients on the structures and statistics, both in the wall and wake regions are discussed.

Spalart, P. R.↗

NASA NTRS1984

Numerical simulation of boundary-layer transition

The transition to turbulence in boundary layers was investigated by direct numerical solution of the nonlinear, three-dimensional, incompressible Navier-Stokes equations in the half-infinite domain over a flat plate. Periodicity was imposed in the streamwise and spanwise directions. A body force was applied to approximate the effect of a nonparallel mean flow. The numerical method was spectra, based on Fourier series and Jacobi polynomials, and used divergence-free basis functions. Extremely rapid convergence was obtained when solving the linear Orr-Sommerfeld equation. The early nonlinear and three-dimensional stages of transition, in a boundary layer disturbed by a vibrating ribbon, were successfully simulated. Excellent qualitative agreement was observed with either experiments or weakly nonlinear theories. In particular, the breakdown pattern was staggered or nonstaggered depending on the disturbance amplitude.

Spalart, P. R.↗

NASA NTRS1984

Two recent extensions of the vortex method

The vortex method, coupled to an integral boundary-layer solver, is applied to the numerical simulation of high-Reynolds-number, separated flows in two new cases: a bluff body in a wind tunnel and flow in a cascade. In the bluff body case, the blockage effect of the tunnel walls is included approximately, assuming an inviscid boundary condition at the walls. The resulting increase in drag is computed, and compares well with a small-disturbance theory and with experiments. The results for flow in a cascade are compared with results from two finite-difference codes for a single-blade case, and good agreement is found. When a staggered cascade is treated with five independent blades, the simulation predicts rotating stall, depending on the angle of attack, and the essential features of the flow are correct. The sensitivity of this phenomenon to various parameters is studied and the stall boundary is found.

Spalart, P. R.↗

NASA NTRS1983

Airloads on bluff bodies, with application to the rotor-induced downloads on tilt-rotor aircraft

The aerodynamic characteristics of airfoils with several flap configurations were studied theoretically and experimentally in environments that simulate a wing immersed in the downwash of a hovering rotor. Special techniques were developed for correcting and validating the wind tunnel data for large blockage effects, and the test results were used to evaluate two modern blockage effects, and the test results were used to evaluate two modern computational aerodynamics codes. The combined computed and measured results show that improved flap and leading-edge configurations can be designed which will achieve large reductions in the downloads of tilt-rotor aircraft, and thereby improve their hover efficiency.

Mccroskey, W. J.↗

NASA NTRS1983

Numerical simulation of vortex breakdown by the vortex-filament method

The vortex filament method was applied to the simulation of vortex breakdown. The principal vortex region was represented by multiple filaments, and an axial velocity component was induced by a spiral winding of the filaments. First, an accuracy check was performed for a cylindrical swirling flow with simple analytical expressions for the axial and theta velocities. The result suggests that the flow field is simulated to any accuracy by increasing the number of filaments. Second, an axisymmetric type vortex breakdown was simulated, with experimental data serving as upstream conditions. The calculated axial and theta velocity contours show the breakdown of the vortex, including a rapid change in the vortex core, followed axially by a recovery zone and then a second breakdown. When three dimensional initial data are used the second breakdown appears to be of the spiral type in correspondence with experimental observations. The present method is easily used to simulate other types of vortex breakdown or other vortex flows with axial velocity.

Nakamura, Y.↗

NASA NTRS1983

Numerical simulation of vortex breakdown by the vortex-filament method

The vortex-filament method was applied to the simulation of vortex breakdown. The principal vortex region was represented by multiple filaments, and an axial velocity component was induced by a spiral winding of the filaments. An accuracy check was performed for a cylindrical swirling flow with simple analytical expressions for the axial and theta velocities. The result suggests that the flow field can be simulated to any accuracy by increasing the number of filaments. An axisymmetric-type vortex breakdown was simulated, with experimental data serving as upstream conditions. The calculated axial- and theta-velocity contours show the breakdown of the vortex, including a rapid change in the vortex core, followed axially by a recovery zone and then a second breakdown. When three dimensional initial data are used the second breakdown appears to be of the spiral type in correspondence with experimental observations. The present method can easily be used to simulate other types of vortex breakdown or other vortex flows with axial velocity.

Nakamura, Y.↗

NASA NTRS1983

Numerical simulation of separated flows

A new numerical method, based on the Vortex Method, for the simulation of two-dimensional separated flows, was developed and tested on a wide range of gases. The fluid is incompressible and the Reynolds number is high. A rigorous analytical basis for the representation of the Navier-Stokes equation in terms of the vorticity is used. An equation for the control of circulation around each body is included. An inviscid outer flow (computed by the Vortex Method) was coupled with a viscous boundary layer flow (computed by an Eulerian method). This version of the Vortex Method treats bodies of arbitrary shape, and accurately computes the pressure and shear stress at the solid boundary. These two quantities reflect the structure of the boundary layer. Several versions of the method are presented and applied to various problems, most of which have massive separation. Comparison of its results with other results, generally experimental, demonstrates the reliability and the general accuracy of the new method, with little dependence on empirical parameters. Many of the complex features of the flow past a circular cylinder, over a wide range of Reynolds numbers, are correctly reproduced.

Spalart, P. R.↗

NASA NTRS1982

Vortex simulation of an inviscid shear layer

The accuracy of the vortex-blob method was tested by simulating a free-shear-layer instability, Kirchhoff's elliptical vortex, and a circular vortex. The main numerical parameters in the vortex-blob method are the density of the vortices, and the distribution of vorticity within each vortex core. The growth rate of a periodic unstable mode of the shear layer was calculated numerically and compared with the exact result. The error is only a few percent for about 10 rows of vortex blobs. The error is reduced by decreasing the spacing between vortices and, correspondingly, the core size. In the simulation of the motion of the elliptical vortex, the rotation of the boundary, without change of shape, and the circular particle paths of the vortical fluid were well simulated. For the circular vortex, optimum sets of parameters were obtained by comparing them with the exact velocity. The results are consistent with convergence theories of the vortex-blob method. In particular, second-order convergence is observed with a Gaussian core from velocity calculation.

Nakamura, Y.↗

NASA NTRS1981

Computation of separated flows by a vortex-tracing algorithm

Numerical solutions for two-dimensional, time-dependent, separated flows around bodies are obtained, using a new version of the vortex method. This method provides an efficient representation of flows involving large regions of separation. The modifications incorporated in the new version, which improve its accuracy, versatility, and computing speed, are described. The computer cost is only of the order of the 3/2 power on N, instead of N-squared, for each step with N vortices. Arbitrary shapes can be treated; a conformal mapping is not required. Special attention is paid to the viscous character of the solution and to the accurate computation of the pressure distribution at the body surface. The vortex solution for the outer flow is coupled to an inner solution for the attached part of the boundary layer. Numerical results are presented for several bluff bodies exhibiting dependence on Reynolds number, for stationary airfoils under steady or transient conditions and for oscillating airfoils, including dynamic stalls. These results are compared with other available results, analytical or experimental, and demonstrate the enhanced reliability and accuracy of the improved method.

Spalart, P. R.↗