The control of turbulent end-wall boundary layers using surface suction

1999 ◽  
Vol 27 (6) ◽  
pp. 484-496 ◽  
Author(s):  
C. V. Seal ◽  
C. R. Smith
1971 ◽  
Vol 93 (2) ◽  
pp. 300-314 ◽  
Author(s):  
G. L. Mellor ◽  
G. M. Wood

The essential ingredient missing in existing prediction methods for the performance of multistage axial compressors is that which would account for the effect of end-wall boundary layers. It is, in fact, believed that end-wall boundary layers play a major role in compressor performance and the absence of an adequate theory represents a handicap to turbomachinery designers that might be likened to the handicap that designers of wings, for example, would face if Prandtl had not introduced the idea of a boundary layer. In this paper a new theory is developed which retains all elements of classical boundary layer theory; for example, we discuss variables such as momentum thickness and wall shear stress. However, the present theory introduces new concepts such as axial and tangential defect force thickness, a rotor exit-stator inlet “jump condition” and the importance of these concepts is demonstrated. Inherent in the derivation is an identification of the role of secondary flow and tip clearance flow. A proper means of matching the boundary layer calculations to conventional main stream calculations is suggested. Independent of empirical parametization it appears that the theory is capable of correctly modeling boundary layer blockage, losses, and end-wall stall. Near stall, the main stream-boundary layer interaction is very strong.


1982 ◽  
Vol 104 (2) ◽  
pp. 467-478 ◽  
Author(s):  
B. Lakshminarayana ◽  
A. Ravindranath

This paper reports the experimental study of the three-dimensional characteristics of the mean velocity of the rotor wake inside the annulus- and hub-wall boundary layers. The measurements were taken with a rotating three-sensor hot wire behind the rotor. This set of measurements probably represents the first set of comprehensive measurements taken inside the annulus- and hub-wall boundary layers. The wake was surveyed at several radial locations inside the boundary layer region and at several axial locations. Interaction of the wake with the annulus-wall boundary layer, secondary flow, tip-leakage flow, and the trailing vortex system results in slower decay and larger width of the wake. The presence of a strong vortex and its merger with the wake is also observed. The end-wall boundary layers and the secondary flow were found to have a substantial effect on both the decay characteristics and the profile of the wake. These and other measurements are reported and interpreted in this paper.


1982 ◽  
Author(s):  
H. Ekerol ◽  
J. W. Railly

Experimental data on the wall shear stress of a turbulent boundary layer on the suction side of a blade in a two-dimensional radial impeller is compared with the predictions of a theory which takes account of rotation and curvature effects as well as the three-dimensional influence of the end-wall boundary layers. The latter influence is assumed to arise mainly from mainstream distortion due to secondary flows created by the end-wall boundary layers and it appears as an extra term in the momentum integral equation of the blade boundary layer which has allowance, also for the Coriolis effect; an appropriate form of the Head entrainment equation is derived to obtain a solution and a comparison made. A comparison of the above theory with the Patankar-Spalding prediction method, modified to include the effects of Coriolis (including mixing length modification, MLM) is also made.


Author(s):  
J Dunham

This paper introduces a new phenomenological approach to modelling the end-wall effects. It is based on explicit calculations of the annulus wall boundary layers by Hirsch and de Ruyck's method, with significant changes to allow tip clearance effects to be represented, together with explicit calculations of the secondary flows outside the boundary layers using Marsh's equations. To achieve this, the secondary vorticity is computed throughout the compressor. A new simple model for the tip clearance vorticity is added. It is shown, by comparison with Salvage's cascade measurements and Inoue's isolated rotor measurements at varying tip clearance, that the method is capable of satisfactory predictions of the pitchwise-average local deviation and loss near the end-walls. The model is only accurate, however, when there is no significant end-wall flow separation. Two examples of complete multistage low-speed compressors are given. It is concluded that a promising foundation has been presented for a more satisfactory and more accurate way of predicting the end-wall effects than any published purely empirical scheme.


Author(s):  
LuCheng Ji ◽  
WeiWei Shao ◽  
WeiLin Yi ◽  
Jiang Chen

This paper presents a model for describing the influences of SUC-EW dihedral angle on corner separation in turbomachinery, in which SUC-EW dihedral angle refers to the dihedral angle at the intersection line between blade ‘SUCtion’ and End-Wall surfaces. Based on the physical intuition of that the three-dimensional (3D) corner boundary layer is the conflux of both blade and end wall boundary layers, an equivalent two-dimensional(2D) corner boundary layer is put forward to predict the behavior of corner boundary layer. In this procedure, the cross flow effect in corner boundary layer and the three-dimensionality of the nearby main flow are ignored. The influence of the SUC-EW dihedral angle is included by another assumption. That is, the aero blockage and momentum loss of both blade and end wall boundary layers are conserved during the procedure of superimposing the two (both blade and end wall) 2D boundary layers to form the equivalent corner one. Then the corner separation is judged by combining the behaviors of the three boundary layers, i.e. the blade, the end wall and the equivalent 2D corner boundary layers. The present model reveals the influence of the SUC-EW dihedral angle and its streamwise gradient on the corner separation. Carefully monitoring and controlling this dihedral angle and its streamwise gradient are important ways to alleviate or even eliminate the corner separation. Simple numerical investigations show that the model is qualitatively correct.


1981 ◽  
Vol 103 (1) ◽  
pp. 20-33 ◽  
Author(s):  
J. De Ruyck ◽  
C. Hirsch

A previously developed axial compressor end-wall boundary layer calculation method which requires the introduction of three-dimensional velocity profile models is summarized. In this method the classical three-dimensional velocity profile models were shown to present inherent limitations at stall limit, with regard to the range of transverse boundary layer thicknesses they are able to represent. A corrected profile model is presented which contains no more limitations without affecting the previous found overall results. Stall limit is predicted by limiting values of shape factor and/or diffusion factor. The new profile model containing also compressibility effects allows the calculation of boundary layers in machines with shrouded blades, by simulating the jump between rotating and non rotating parts of the walls. A corrected version of a force defect correlation is presented which is shown to give better agreement at high incidences. Some results on high and low speed machines are discussed. The model is applied to obtain an end-wall blockage correlation depending on geometry, flow coefficient, AVR, aspect ratio, solidity, diffusion factor, Reynolds number, axial blade spacing, tip clearance and inlet boundary layer thickness. A quantitative estimation of the losses associated with the end-wall boundary layers can be obtained using this analysis and therefore can be a useful tool in the design of an axial compressor stage.


Sign in / Sign up

Export Citation Format

Share Document