Computation of viscous flows in turbomachinery cascades with a space-marching method
Explore the source record for details and available documents.
Pouagare, M.↗
Engineering topics
Pouagare, M.
Publications and source records attributed to Pouagare, M..
A space-marching method for viscous incompressible internal flows
A numerical algorithm for calculating steady subsonic two-dimensional incompressible channel flows in ducts is developed and extended to three-dimensional flows. A space-marching method similar to that developed by Schiff and Steger (1980) for supersonic flows is applied, and numerical results for developing laminar flows in a two-dimensional channel, in a straight square duct, and in a mildly curved square duct are presented graphically. Good agreement with experimental data and analytical results is found.
Pouagare, M.↗
A space-marching method for incompressible Navier-Stokes equations
This paper deals with the development of a space-marching method for incompressible flows. The method solves the continuity and momentum equations as a coupled system at each streamwise station. The character of the system of equations has been changed from elliptic to hyperbolic/parabolic in order to enable the equations to be marched in space. The present method has many advantages compared to the existing parabolic or space-marching methods for incompressible flow: (1) it avoids the solution of Poisson equations, (2) it conserves the mass flow with no additional computation, (3) it does not require the specification of an assumed pressure field when used in the prediction of duct flows. The present method can capture strong secondary velocities and strong transverse pressure gradients. Predictions of the flow through straight and curved ducts are in good agreement with analytical and experimental results.
Pouagare, M.↗
An experimental study of the compressor rotor blade boundary layer
The three-dimensional turbulent boundary layer developing on a rotor blade of an axial flow compressor was measured using a miniature 'x' configuration hot-wire probe. The measurements were carried out at nine radial locations on both surfaces of the blade at various chordwise locations. The data derived includes streamwise and radial mean velocities and turbulence intensities. The validity of conventional velocity profiles such as the 'power law profile' for the streamwise profile, and Mager and Eichelbrenner's for the radial profile, is examined. A modification to Mager's crossflow profile is proposed. Away from the blade tip, the streamwise component of the blade boundary layer seems to be mainly influenced by the streamwise pressure gradient. Near the tip of the blade, the behavior of the blade boundary layer is affected by the tip leakage flow and the annulus wall boundary layer. The 'tangential blockage' due to the blade boundary layer is derived from the data. The profile losses are found to be less than that of an equivalent cascade, except in the tip region of the blade.
Pouagare, M.↗
Computation and turbulence closure models for shear flows in rotating curved bodies
This paper is concerned with the effects of rotation and curvature on the turbulence closure model. The k-epsilon model that includes the effects of curvature has been modified to include the anisotropy in turbulence. A space-marching algorithm which solves the parabolized Navier-Stokes equations is coupled with the turbulence closure model to predict the turbulent shear flow over a rotating cylinder and inside a rotating channel. The agreement between the predictions and available data is good. The rotation and curvature effects are predicted well.
Pouagare, M.↗
Three-dimensional flowfield inside a low-speed axial flow compressor rotor
Previously cited in issue 15, p. 2347, Accession no. A82-31964
Pouagare, M.↗
Annulus wall boundary layer development in a compressor stage, including the effects of tip clearance
The end-wall boundary layer development in a compressor stage, including the inlet guide vane (IGV) passage and the rotor passage, was measured. The measurement upstream of the rotor and inside the IGV passage were carried out with a five-hole probe. The data (blade-to-blade) inside the IGV passage were carried out with a five-hole probe. The data (blade-to-blade) inside the rotor passage were measured using a three-sensor rotating hot-wire below the tip clearance region and "V' configuration probe inside the clearance region. The rotor exit measurements (blade-to-blade) were acquired with a laser Doppler velocimeter. The velocity profiles and the integral properties are presented and interpreted. The boundary layer is comparatively well behaved up to the leading edge of the rotor, beyond which complex interactions result in very unconventional profiles. The momentum thicknesses decrease in the leakage flow region of the rotor. The momentum thicknesses and the limiting streamline angles predicted from a momentum integral technique agree well with the data up to the leading edge of the rotor.
Lakshminarayana, B.↗
Three dimensional flow field inside compressor rotor, including blade boundary layers
The space marching code was modified in order to be able to predict the flow field inside a rotor passage, including the blade and hub wall boundary layers. The basic changes incorporated are modifications of the equations so that the code can handle three dimensional configurations with changes in the radial direction (for example changes in stagger angle, blade camber and thickness), extensions and modifications in order to implement a physically realistic turbulence model such as a k sigma model and an algebraic Reynolds stress model.
Pouagare, M.↗
Computation of viscous flows in turbomachinery cascades
Analytic and numerical experiments involving the use of the space-marching method for the prediction of the viscous flow field in turbomachinery cascades indicate that the grid lines in the physical domain must be nearly orthogonal in order to arrive at accurate predictions. For staggered cascades, a nonperiodic grid system was used and appropriate approximations were applied in place of the periodicity boundary conditions upstream and downstream of the cascade. Predicted drag coefficient, turning angle, boundary layer momentum thicknesses, and velocity profiles are compared with experimental data. Good agreement is obtained in most cases.
Pouagare, M.↗
Development of secondary flow and vorticity in curved ducts, cascades, and rotors, including effects of viscosity and rotation
This paper is concerned with the numerical solution of the secondary vorticity equations in curved ducts, cascades, and rotors. The classical approach of splitting the flow into primary and secondary flow fields is employed and extended to include effects of viscosity, rotation, and density stratification. All elliptical effects are neglected and the Crank-Nicolson method is used to solve the secondary vorticity equation. The secondary flow field is obtained by solving a Poisson equation using a successive over-relaxation method. The results are compared with theoretical and experimental data from stationary ducts, compressor and turbine cascades. The agreement is good for most of the cases.
Pouagare, M.↗
Three dimensional flow field inside the passage of a low speed axial flow compressor rotor
Measurements of the subsonic flow in the rotor passage of a single stage axial flow compressor were made to study the nature of the flow field and to verify the existing numerical codes. The velocity and pressure fields were measured across the entire rotor passage at six axial locations and at five radial locations. A five-hole probe, rotating with the rotor, was used to measure the three components of velocity, the static and the total pressure. The experimental results are compared with the predictions from Katsanis and McNally's computer program. The agreement between the two is good for most of the cases.
Pouagare, M.↗
Three-dimensional flow field in the tip region of a compressor rotor passage. I - Mean velocity profiles and annulus wall boundary layer
A rotating three-sensor, hot-wire probe has been used along with rotor blade static pressure measurements to investigate the complex inviscid and viscid effects in the annulus wall flowfield, including the three components of mean velocity, turbulence intensity, and turbulence stress inside the rotor blade passage. It is found that the tip leakage flow originates near quarter-chord, with peak values occurring near mid-chord. The leakage flow, which is in the form of a jet within the blade row, is augmented by the blade rotation and travels further away from the suction surface than that observed in stationary blade rows and cascades. This leakage flow tends to roll up between the mid-passage and the pressure surface near the tip region; the vortex formation does not occur within the passage in this particular case.
Lakshminarayana, B.↗
Three-dimensional flow field in the tip region of a compressor rotor passage. II - Turbulence properties
The turbulence properties in the annulus wall region of an axial flow compressor rotor was measured using a triaxial, hot-wire probe rotating with the rotor. The flow was surveyed across the entire passage at five axial locations (leading edge, 1/4 chord, 1/2 chord, 3/4 chord, and the trailing edge location) and at six radial locations in a low-speed compressor rotor. The data derived include all three components of turbulence intensity and three components of turbulence stress. A comprehensive interpretation of the data with emphasis on features related to rotation, leakage flow, annulus wall boundary layer, and blade boundary layer interactions is included. All the components of turbulent intensities and stresses are found to be high in the leakage-flow mixing region. The radial component of intensities and stresses is found to be much higher than the corresponding streamwise components. The turbulent spectra clearly reveal the decay process of the inlet-guide-vane wake within the rotor passage.
Lakshminarayana, B.↗
Three dimensional flow field inside compressor rotor, including blade boundary layers
The flow in a turbomachinery blade passage has a predominant flow direction. The viscous diffusion in the streamwise direction is usually small and the elliptic influence is transmitted upstream through the pressure field. Starting with a guessed pressure field, it is possible to converge on the full elliptic solution by iterating between a parabolic solution and an iteration of the pressure field. The main steps of the calculation are given. The blade boundary layers which are three dimensional with laminar, transitional, turbulent, and separation zones are investigated. The kinetic energy is analyzed, and the dissipation equation is presented. Measurements were made of the three dimensional flow inside an axial flow compressor passage.
Pouagare, M.↗
End wall flows in rotors and stators of a single stage compressor
A solution of the flow in the rotor end wall region, including the effects of tip clearance flow, is presented. A method for leakage flow measurement at the tip of a compressor rotor blade is discussed. Measurements are given for a rotor hub wall boundary layer.
Govindam, T. R.↗