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.
The use of a fast elliptic solver in combination with relaxation is presented as an effective way to accelerate the convergence of transonic flow calculations, particularly when a marching scheme can be used to treat the supersonic zone in the relaxation process.
A new computer program is presented for calculating the quasi-steady transonic flow past a helicopter rotor blade in hover as well as in forward flight. The program is based on the full potential equations in a blade attached frame of reference and is capable of treating a very general class of rotor blade geometries. Computed results show good agreement with available experimental data for both straight and swept tip blade geometries.
The theoretical and numerical bases of a program for the solution of the Euler equations with viscous-inviscid interaction for high Reynolds number transonic flow past wing/body configurations are explained. The emphasis is upon the logic behind the equation development. The program is fully detailed so that the user can quickly become familiar with its operation.
The method of matched asymptotic expansions is used in analyzing the structure of the interaction region formed when a shock wave impinges on a turbulent flat plate boundary layer in transonic flow. Solutions in outer regions, governed by inviscid flow equations, lead to relations for the wall pressure distribution. Solutions in the inner regions, governed by equations in which Reynolds and/or viscous stresses are included, lead to a relation for the wall shear stress. Solutions for the wall pressure distribution are reviewed for both oblique and normal incoming shock waves. Solutions for the wall shear stress are discussed.
A finite-difference relaxation method is presented for numerical solution of the full potential equation and exact boundary conditions for general axisymmetric bodies is inviscid, steady transonic flow. Body-normal coordinates are used in the nose region and sheared cylindrical coordinates are used on the afterbody to accommodate corners such as boattails and flares. An improved difference scheme is used which does not require that the flow be nearly alined with a coordinate direction in supersonic regions, and which treats either subsonic or supersonic free streams. Numerical results are illustrated for some simple classical shapes such as spheres and ellipsoids, and for more practical shapes like tangent-ogives with boattails. Special attention is given to bodies which have been studied for area-rule applications. Agreement with available experimental results is good in cases where viscous effects and wind-tunnel wall interference are not important.
The results reported here are based on applying the method of parametric differentiation (MPD) to transform the nonlinear differential equation governing small-disturbance transonic flow to a linear equation. Implicit approximate factorization and monotone methods were used to accelerate convergence of the linear problem by an order of magnitude over successive line over-relaxation. The relative merits of using MPD are discussed in comparison to conventional small-disturbance applications. Several MPD analyses are performed on an array of airfoils. A design procedure utilizing MPD is discussed and demonstrated for two nonlifting cases.
The influence of multiple high pressure, supersonic, radial or tangential jets, that are injected from the circumference of the base plane of an axisymmetric body, on its longitudinal aerodynamic coefficients in transonic flow is studied experimentally. The interaction of the jets with the body flow field increases the pressures on the forebody, thus altering its lift and static stability characteristics. It is shown that, within the range of parameters studied. This interaction has a stabilizing effect on the body. The contribution to lift and stability is significant at small angles of attack and decreases nonlinearly at higher angles when the crossflow mechanism becomes dominant.
An implicit delta form finite-difference algorithm for Euler equations in conservation law form has been used in preliminary calculations of three-dimensional wing-vortex interactions. Both steady and unsteady transonic flow wing-vortex interactions are computed. The computations themselves are meant to guide upcoming wind tunnel experiments of the same flow field. Various modifications to the numerical method that are intended to improve computational efficiency are also described and tested in both two- and three-dimensions.
An implicit delta form finite difference algorithm for Euler equations in conservation law form was used in preliminary calculations of three dimensional wing vortex interaction. Both steady and unsteady transonic flow wing vortex interactions are computed. The computations themselves are meant to guide upcoming wind tunnel experiments of the same flow field. Various modifications to the numerical method that are intended to improve computational efficiency are also described and tested in both two and three dimensions. Combination of these methods can reduce the overall computational time by a factor of 4.
A numerical study of transverse curvature effects on compressible flow temporal stability for transonic to low supersonic Mach numbers is presented for axisymmetric modes. The mean flows studied include a similar boundary-layer profile and a nonsimilar axisymmetric boundary-layer solution. The effect of neglecting curvature in the mean flow produces only small quantitative changes in the disturbance growth rate. For transonic Mach numbers (1-1.4) and aerodynamically relevant Reynolds numbers (5000-10,000 based on displacement thickness), the maximum growth rate is found to increase with curvature - the maximum occurring at a nondimensional radius (based on displacement thickness) between 30 and 100.
Various nonreflecting far-field boundary condition procedures are compared by implementing them in the computer code LTRAN2. This code solves the implicit finite-difference representation of the small-disturbance equations for transonic flows about airfoils. The first- and second-approximate nonreflecting conditions, as proposed by Engquist and Majda, are compared with the condition derived from the full-characteristic equation. The far-field boundary conditions and the description of the algorithm for implementing these conditions in LTRAN2 are discussed. Various cases are computed and compared with results from the older, more conventional procedures. One concludes that the full-characteristic equation produces the most effective results, thus allowing the far-field boundary to be located closer to the airfoil; this decreases the computer time required to obtain the solution because fewer mesh points are required.
Three different simulation approaches, namely unsteady Reynolds-averaged Navier-Stokes (URANS), delayed detached-eddy simulation (DDES),and wall-modeled large-eddy simulation (WMLES)are employed to simulate transonic flow over an NACA-0012 airfoil at different angles of attack covering pre- and post-buffet-onset regimes. The freestream Mach number is 0.75,and the Reynolds number based on the chord length is 10million. These conditions are the same as the wind-tunnel experimental conditions of McDevitt and Okuno (1985).1The NASA FUN3D solver is used for the simulations, which is an unstructured, compressible flow solver. The URANS simulations are performed using the Spalart-Allmaras (SA) model with the compressibility correction, the DDES predictions are based on the SA model, and the WMLES are performed using an equilibrium wall-model.The unsteady RANS simulations,only with the compressibility correction,predict the pre- and post-buffet characteristics,which compare well with the experimental results. DDES results predicted a lower buffet onset angle compared to experiment.The predicted shock locations are upstream of the locations predicted by URANS. Using a fine grid in the spanwise direction, WMLES predictions show buffeting consistent with the experiment
Three different simulation approaches, namely unsteady Reynolds-averaged Navier-Stokes (URANS), delayed detached-eddy simulation (DDES), and wall-modeled large-eddy simulation (WMLES) are employed to simulate transonic flow over an NACA-0012 airfoil at different angles of attack covering pre- and post-buffet-onset regimes. The freestream Mach number is 0.75, and the Reynolds number based on the chord length is 10 million. These conditions are the same as the wind-tunnel experimental conditions of McDevitt and Okuno (1985). The NASA FUN3D solver is used for the simulations, which is an unstructured, compressible flow solver. The URANS simulations are performed using the Spalart-Allmaras (SA) model with the compressibility correction, the DDES predictions are based on the SA model, and the WMLES are performed using an equilibrium wall-model. The unsteady RANS simulations, only with the compressibility correction, predict the pre- and post- buffet characteristics, which compare well with the experimental results. DDES results predicted a lower buffet onset angle compared to experiment. The predicted shock locations are upstream of the locations predicted by URANS. Using a fine grid in the spanwise direction, WMLES predictions show buffeting consistent with the experiment.
Three different simulation approaches, namely unsteady Reynolds-averaged Navier-Stokes (URANS), delayed detached-eddy simulation (DDES), and wall-modeled large-eddy simulation (WMLES) are employed to simulate transonic flow over an NACA-0012 airfoil at different angles of attack covering pre-and post-buffet-onset regimes. The freestream Mach number is 0.75,and the Reynolds number based on the chord length is 10million. These conditions are the same as the wind-tunnel experimental conditions of McDevitt and Okuno (1985). The NASA FUN3D solver is used for the simulations, which is an unstructured, compressible flow solver. The URANS simulations are performed using the Spalart-Allmaras (SA) model with the compressibility correction, the DDES predictions are based on the SA model, and the WMLES are performed using an equilibrium wall-model. The unsteady RANS simulations, only with the compressibility correction, predict the pre-and post-buffet characteristics, which compare well with the experimental results. DDES results predicted a lower buffet onset angle compared to experiment. The predicted shock locations are upstream of the locations predicted by URANS. Using a fine grid in the span wise direction, WMLES predictions show buffeting consistent with the experiment.
Thin layer Navier-Stokes and Euler equations are numerically solved using a multi-block zonal approach to simulate the formation and roll up of tip vortices of wings in subsonic and transonic flows. Several wing planforms were considered to examine the influence of tip-cap shape, planform geometry and free stream Mach number on the formation process. A good definition of the formation and qualitative roll up of tip vortices was achieved.
A new algorithm was developed for the Euler and Navier-Stokes equations that uses upwind differencing based on the streawise direction. This algorithm is time accurate and can be used in codes for calculating unsteady transonic flows over wings. Such codes can be used for the flutter analysis of wings. In this algorithm, the coordinate system is locally rotated to align with the streamwise direction. For differencing the convective terms in the streamwise direction, a new form of flux splitting is employed, in which the biasing depends on the local Mach number. In the plane perpendicular to the stream direction, the new flux splitting uses the condition of no flow in that local plane. By using a locally rotated coordinate system, the convective flux vector biasing depends on the total Mach number. Hence, the switching of the flux vector biasing occurs across shock waves and the proper domain of dependence is used in supersonic regions. For comparison, many other upwind methods switch differencing based on Mach number of shock waves in multidimensional flows. The formulas for the convective flux vector differencing do not contain any user specified parameters. So, the amount of numerical dissipation is automatically determined.
A new algorithm was developed for the Euler and Navier-Stokes equations that uses upwind differencing based on the streamwise direction. This algorithm is time accurate and can be used in codes for calculating unsteady transonic flows over wings. Such codes can be used for the flutter analysis of wings. In this algorithm, the coordinate system is locally rotated to align with the streamwise direction. For differencing the convective terms in the streamwise direction, a new form of flux splitting is employed, in which the biasing depends on the local Mach number. In the plane perpendicular to the stream direction, the new flux splitting uses the condition of no flow in that local plane. By using a locally rotated coordinate system, the convective flux vector biasing depends on the total Mach number. Hence, the switching of the flux vector biasing occurs across shock waves and the proper domain of dependence is used in supersonic regions. For comparison, many other upwind methods switch differencing based on Mach number of shock wave in multidimensional flows. The formulas for the convective flux vector differencing do not contain any user specified parameters. So, the amount of numerical dissipation is automatically determined.
An advanced laser anemometer (LA) was used to measure the axial and tangential velocity components in an annular cascade of turbine stator vanes operating at transonic flow conditions. The vanes tested were based on a previous redesign of the first-stage stator in a two-stage turbine for a high-bypass-ratio engine. The vanes produced 75 deg of flow turning. Tests were conducted on a 0.771-scale model of the engine-sized stator. The advanced LA fringe system employed an extremely small 50-micron diameter probe volume. Window correction optics were used to ensure that the laser beams did not uncross in passing through the curved optical access port. Experimental LA measurements of velocity and turbulence were obtained at the mean radius upstream of, within, and downstream of the stator vane row at an exit critical velocity ratio of 1.050 at the hub. Static pressures were also measured on the vane surface. The measurements are compared, where possible, with calculations from a three-dimensional inviscid flow analysis. Comparisons were also made with the results obtained previously when these same vanes were tested at the design exit critical velocity ratio of 0.896 at the hub. The data are presented in both graphical and tabulated form so that they can be readily compared against other turbomachinery computations.