A Geometry Generation Framework for Contoured Endwalls

Author(s):  
L. E. Wood ◽  
R. R. Jones ◽  
O. J. Pountney ◽  
J. A. Scobie ◽  
D. A. S. Rees ◽  
...  

Abstract The mainstream, or primary, flow in a gas turbine annulus is characteristically two-dimensional over the mid-span region of the blading, where the radial flow is almost negligible. Contrastingly, the flow in the endwall and tip regions of the blading is highly three-dimensional, characterised by boundary layer effects, secondary flow features and interaction with cooling flows. Engine designers employ geometric contouring of the endwall region in order to reduce secondary flow effects and subsequently minimise their contribution to aerodynamic loss. Such is the geometric variation of vane and blade profiles — which has become a proprietary art form — the specification of an effective endwall geometry is equally unique to each blade-row. Endwall design methods, which are often directly coupled to aerodynamic optimisers, are widely developed to assist with the generation of contoured surfaces. Most of these construction methods are limited to the blade-row under investigation, while few demonstrate the controllability required to offer a universal platform for endwall design. This paper presents a Geometry Generation Framework (GGF) for the generation of contoured endwalls. The framework employs an adaptable meshing strategy, capable of being applied to any vane or blade, and a versatile function-based approach to defining the endwall shape. The flexibility of this novel approach is demonstrated by recreating a selection of endwalls from the literature, which were selected for their wide-range of contouring approaches.

2019 ◽  
Vol 142 (1) ◽  
Author(s):  
Liam E. Wood ◽  
Robin R. Jones ◽  
Oliver J. Pountney ◽  
James A. Scobie ◽  
D. Andrew S. Rees ◽  
...  

Abstract The mainstream, or primary, flow in a gas turbine annulus is characteristically two-dimensional over the midspan region of the blading, where the radial flow is almost negligible. Contrastingly, the flow in the endwall and tip regions of the blading is highly three-dimensional (3D), characterized by boundary layer effects, secondary flow features, and interaction with cooling flows. Engine designers employ geometric contouring of the endwall region in order to reduce secondary flow effects and subsequently minimize their contribution to aerodynamic loss. Such is the geometric variation of vane and blade profiles—which has become a proprietary art form—the specification of an effective endwall geometry is equally unique to each blade row. Endwall design methods, which are often directly coupled to aerodynamic optimizers, are widely developed to assist with the generation of contoured surfaces. Most of these construction methods are limited to the blade row under investigation, while few demonstrate the controllability required to offer a universal platform for endwall design. This paper presents a geometry generation framework (GGF) for the generation of contoured endwalls. The framework employs an adaptable meshing strategy, capable of being applied to any vane or blade, and a versatile function-based approach to defining the endwall shape. The flexibility of this novel approach is demonstrated by recreating a selection of endwalls from the literature, which were selected for their wide range of contouring approaches.


Author(s):  
C. Xu ◽  
R. S. Amano ◽  
B. Marini

A three-dimensional computational code was developed for solving time-averaged flows within a turbine blade row using a novel time-marching method. A concept of incorporating dissipation terms into the time derivative terms was proposed to allow the code to have the capability of handling both incompressible and compressible flows. The code was validated by comparing the computational results with experiments in a turbine stator blade passage. The code was further used to investigate the influence of secondary flow in a turbine blade row due to different pitch-width ratios. Detailed secondary flows as well as loss profiles in different sizes of root pitch-width ratio are presented and discussed. The results of this study provide useful information for evaluation of the secondary flow effects due to the pitch-width ratio influence for the future new turbine blade designs.


2006 ◽  
Vol 129 (2) ◽  
pp. 212-220 ◽  
Author(s):  
Giovanna Barigozzi ◽  
Giuseppe Franchini ◽  
Antonio Perdichizzi

The present paper reports on the aerothermal performance of a nozzle vane cascade, with film-cooled end walls. The coolant is injected through four rows of cylindrical holes with conical expanded exits. Two end-wall geometries with different area ratios have been compared. Tests have been carried out at low speed (M=0.2), with coolant to mainstream mass flow ratio varied in the range 0.5–2.5%. Secondary flow assessment has been performed through three-dimensional (3D) aerodynamic measurements, by means of a miniaturized five-hole probe. Adiabatic effectiveness distributions have been determined by using the wide-band thermochromic liquid crystals technique. For both configurations and for all the blowing conditions, the coolant share among the four rows has been determined. The aerothermal performances of the cooled vane have been analyzed on the basis of secondary flow effects and laterally averaged effectiveness distributions; this analysis was carried out for different coolant mass flow ratios. It was found that the smaller area ratio provides better results in terms of 3D losses and secondary flow effects; the reason is that the higher momentum of the coolant flow is going to better reduce the secondary flow development. The increase of the fan-shaped hole area ratio gives rise to a better coolant lateral spreading, but appreciable improvements of the adiabatic effectiveness were detected only in some regions and for large injection rates.


Author(s):  
S. Friedrichs ◽  
H. P. Hodson ◽  
W. N. Dawes

The endwall film-cooling cooling configuration investigated by Friedrichs et al. (1996, 1997) had in principle sufficient cooling flow for the endwall, but in practice, the redistribution of this coolant by secondary flows left large endwall areas uncooled. This paper describes the attempt to improve upon this datum cooling configuration by redistributing the available coolant to provide a better coolant coverage on the endwall surface, whilst keeping the associated aerodynamic losses small. The design of the new, improved cooling configuration was based on the understanding of endwall film-cooling described by Friedrichs et al. (1996, 1997). Computational fluid dynamics were used to predict the basic flow and pressure field without coolant ejection. Using this as a basis, the above described understanding was used to place cooling holes so that they would provide the necessary cooling coverage at minimal aerodynamic penalty. The simple analytical modelling developed in Friedrichs et al. (1997) was then used to check that the coolant consumption and the increase in aerodynamic loss lay within the limits of the design goal. The improved cooling configuration was tested experimentally in a large scale, low speed linear cascade. An analysis of the results shows that the redesign of the cooling configuration has been successful in achieving an improved coolant coverage with lower aerodynamic losses, whilst using the same amount of coolant as in the datum cooling configuration. The improved cooling configuration has reconfirmed conclusions from Friedrichs et al. (1996, 1997); firstly, coolant ejection downstream of the three-dimensional separation lines on the endwall does not change the secondary flow structures; secondly, placement of holes in regions of high static pressure helps reduce the aerodynamic penalties of platform coolant ejection; finally, taking account of secondary flow can improve the design of endwall film-cooling configurations.


Author(s):  
P W James

The purpose of this paper is, firstly, to show how the concept of excess secondary vorticity arises naturally from attempts to recover three-dimensional flow details lost in passage-averaging the equations governing the flow through gas turbines. An equation for the growth of excess streamwise vorticity is then derived. This equation, which allows for streamwise entropy gradients through a prescribed loss term, could be integrated numerically through a blade-row to provide the excess vorticity at the exit to a blade-row. The second part of the paper concentrates on the approximate methods of Smith (1) and Came and Marsh (2) for estimating this quantity and demonstrates their relationship to each other and to the concept of excess streamwise vorticity. Finally the relevance of the results to the design of blading for gas turbines, from the point of view of secondary flow, is discussed.


2000 ◽  
Vol 123 (2) ◽  
pp. 274-280 ◽  
Author(s):  
Sang Woo Lee ◽  
Seong Kuk Joo ◽  
Joon Sik Lee

Presented are three-dimensional mean velocity components and aerodynamic loss data inside circular injection holes. The holes are normally oriented to a crossflow and each hole has a sharp square-edged inlet. Because of their importance to flow behavior, three different blowing ratios, M=0.5, 1.0, and 2.0, and three hole length-to-diameter ratios, L/D=0.5, 1.0, and 2.0, are investigated. The entry flow is characterized by a separation bubble, and the exit flow is characterized by direct interaction with the crossflow. The uniform oncoming flow at the inlet undergoes a strong acceleration and a subsequent gradual deceleration along a converging–diverging flow passage formed by the inlet separation bubble. After passing the throat of the converging–diverging passage, the potential core flow, which is nearly axisymmetric, decelerates on the windward side, but tends to accelerate on the leeward side. The presence of the crossflow thus reduces the discharge of the injectant on the windward side, but enhances its efflux on the leeward side. This trend is greatly accentuated at M=0.5. In general, there are strong secondary flows in the inlet and exit planes of the injection hole. The secondary flow within the injection hole, on the other hand, is found to be relatively weak. The inlet secondary flow is characterized by a strong inward flow toward the injection-hole center. However, it is not completely directed inward since the crossflow effect is superimposed on it. Past the throat, secondary flow is observed such that the leeward velocity component induced by the crossflow is superimposed on the diverging flow. Short L/D usually results in an exit discharging flow with a steep velocity gradient as well as a strong deceleration on the windward side, as does low M. The aerodynamic loss inside the injection hole originates from the inlet separation bubble, wall friction and interaction of the injectant with the crossflow. The first one is considered as the most dominant source of loss, even in the case of L/D=2.0. At L/D=0.5, the first and third sources are strongly coupled with each other. Regardless of L/D, the mass-averaged aerodynamic loss coefficient has an increasing tendency with increasing M.


Author(s):  
J. D. Denton ◽  
L Xu

Many of the phenomena involved in turbomachinery flow can be understood and predicted on a two-dimensional (2D) or quasi-three-dimensional (Q3D) basis, but some aspects of the flow must be considered as fully three-dimensional (3D) and cannot be understood or predicted by the Q3D approach. Probably the best known of these fully 3D effects is secondary flow, which can only be predicted by a fully 3D calculation which includes the vorticity at inlet to the blade row. It has long been recognized that blade sweep and lean also produce fully 3D effects and approximate methods of calculating these have been developed. However, the advent of fully 3D flow field calculation methods has made predictions of these complex effects much more readily available and accurate so that they are now being exploited in design. This paper will attempt to describe and discuss fully 3D flow effects with particular reference to their use to improve turbomachine performance. Although the discussion is restricted to axial flow machines, many of the phenomena discussed are equally applicable to mixed and radial flow turbines and compressors.


Author(s):  
A. Binder ◽  
R. Romey

In highly loaded turbines with large hub/tip ratios there is a marked increase in secondary flow effects. The optimization of the turbine flow requires detailed knowledge both of three-dimensional cascade flow and of the wake impinging on the downstream rows of airfoils. Therefore, in the DFVLR, thorough investigations of a single-stage turbine with cold air flow were performed. The stator of this turbine was designed for transonic flow and has a hub/tip ratio of 0.756 and an aspect ratio of 0.56. First, measurements were taken without the rotor in several sections behind the turbine stator with special regard to the mixing of the wakes and secondary vortices. Distributions of total pressure loss coefficient and flow direction give the order of magnitude of the mixing losses. Also, position, intensity, structure, and development of secondary vortices are shown. Some complementary measurements were carried out using five-hole probes. They confirm the above described results from two-dimensional measurements.


Author(s):  
J. F. Carrotte ◽  
K. F. Young ◽  
S. J. Stevens

A series of tests have been carried out to investigate the flow in a Compressor Outlet Guide Vane (OGV) blade row downstream of a single stage rotor. The subsequent flow field that developed within an OGV passage was measured, at intervals of 10% axial chord, using a novel design of miniature 5 hole pressure probe. In addition to indicating overall pressure levels and the growth of regions containing low energy fluid, secondary flow features were identified from calculated axial vorticity contours and flow vectors. Close to each casing the development of classical secondary flow was observed, but towards the centre of the annulus large well defined regions of opposite rotation were measured. These latter flows were due to the streamwise vorticity at inlet to the blade row associated with the skewed inlet profile. Surface static pressures were also measured and used to obtain the blade pressure force at 3 spanwise locations. These values were compared with the local changes in flow momentum calculated from the measured velocity distributions. With the exception of the flow close to the outer casing, which is affected by rotor tip leakage, good agreement was found between these quantities indicating relatively weak radial mixing.


1977 ◽  
Vol 99 (1) ◽  
pp. 29-36 ◽  
Author(s):  
J. W. Railly ◽  
P. B. Sharma

Hitherto, theories of annulus wall boundary layer development in axial compressors have assumed an axially-symmetric flow in which the blade action has been replaced by a force field. A more rigorous treatment of the momentum equations in the annulus boundary layer by Mellor and Wood demonstrated the presence of certain terms, after the equations had been averaged in the pitch-wise direction, which arise from the truly three-dimensional character of the flow. These terms, which may be described as the gradients of apparent stresses, were not regarded by them (apart from a discussion of tip clearance) as having importance for the problem. In the present work a second equation of the annulus wall boundary layer is obtained by consideration of the work of these apparent stresses. By integration of the system of equations over a single blade row, two equations are obtained relating various integral quantities at inlet to and exit from the row. Each equation contains terms which depend upon apparent stresses connected with the relative velocity field at the exit plane. An experiment is described in which the six turbulent stresses in the stationary frame downstream of a single rotor, determined by means of a multiple hot wire array, are used to evaluate each term of the aforementioned equations. The integral quantities thus determined are shown to be reasonably consistent with the predictions from the two equations, in particular, for the case of the hub boundary layer. Theoretical solutions of the two integral equations require a secondary flow hypothesis so that the departure from collateral flow at blade row exit is determined by the solution.


Sign in / Sign up

Export Citation Format

Share Document