Proceedings of

2000 International Joint Power Generation Conference

Miami Beach, Florida, July 23-26, 2000

IJPGC2000-15015

THE APPLICATION OF STATOR BLADE COMPOUND LEAN AT ROOT

TO INCREASE THE EFFICIENCY OF LP TURBINE STAGES
FROM LOW TO NOMINAL LOAD
Piotr Lampart
Institute of Fluid Flow Machinery
Polish Academy of Sciences, Gdask, Poland
e-mail: lampart@imppan.imp.pg.gda.pl

turbines, efforts are concentrated on reduction of tip and root leakage,

boundary layer and secondary flow losses. In addition to that in LP
turbines, strong span-wise gradients of reaction, supersonic flow and
shock-related losses are the main concern.
An interesting method of 3D blade stacking is compound lean
where the endwall sections of the stator blades are non-linearly displaced in the circumferential direction relative to mid-span sections.
Investigations of compound lean (at the root and tip) in single HP stator cascades are reported by Harrison, 1992 (experiment), Denton &
Xu, 1999 (theory and CFD). The effects of stator blade compound
lean on the operation of HP turbine stages (stator, subsequent moving
blade row and stage as a whole) are explained by Wang, 1999 (experiment and theory), Singh et al., 1995 (experiment and CFD), Lampart et al. 1999abc (CFD). The univocal conclusion of these works is
that compound lean redistributes blade load, mass flow rate and loss
span-wise. Compound lean has a beneficial effect on reduction of mixing losses and span-wise variations of the mean exit flow angle downstream of the stator trailing edge which is likely to reduce losses in the
blade row downstream as well as make matching easier and improve
off-design performance. The quantitative effect of compound lean on
the overall loss coefficient varies with stage geometry, especially with
the span/chord ratio, and is not very impressive as the efficiency of HP
cylindrical blading is quite high, save for evident failures in
pitch/chord/stagger angle optimisation.
If there are still places where the efficiency of blading systems
can be significantly improved, it is in the LP part, including exit stages
with long blades and diverging contours operating over a wide range
of volumetric flow rates from nominal operating conditions. The present paper is intended to contribute to explain the effect of stator blade
compound lean at the root on the performance of LP turbine stages
and raise our understanding of flow control in LP turbines. The paper
gives CFD-predicted values of efficiency gains that can be achieved
from the tested stage in a wide range of volumetric flow rate.

ABSTRACT
LP turbines are characteristic for their long blades, diverging contours, high velocities and strong radial gradients of pressure, Mach
number and flow angle, especially downstream of the stator at the inlet
to the moving blade. Low reaction at the root can lead to massive
separation zones at the root of rotor blades, which happens especially
at low loads. An interesting design to improve the flow through the
root passage of the overloaded stator and underloaded moving blade
row is compound lean at the root of stator blades. It is a design where
the root sections are nonlinearly displaced in the circumferential direction relative to the mid-span sections. As a result, the leading and trailing edge assume certain curvature span-wise in a plane normal to the
turbine axis. The present paper is meant to describe results of numerical investigations from a 3D NS solver FlowER conducted for several
configurations of stator blade compound lean for different circumferential displacements of the root sections relative to mid-span profiles.
The computations are carried out for a wide range of volumetric
flow rates, accounting for the nominal operating regime as well as low
and high load. It is found that compound lean induces additional blade
force, streamwise curvature and redistribution of flow parameters in
the stage, including pressure and mass flow rate span-wise. This can
improve the flow conditions both in the stator and the rotor. The obtained efficiency improvements depend greatly on the flow regime,
with the highest gains in the region of low load. However, this design
can be unfavourable for high mass flow rates.
INTRODUCTION
With the development of 3D numerical solvers for turbomachinery
flow prediction, improved turbine designs can be sought for with the
aid of computers, not involving experimental costs and risks. One way
to improve the turbine performance is to raise the efficiency of the
blading systems, also by appropriate 3D stacking of the blades. This is
accomplished in different ways for HP and LP turbines. In HP impulse

Copyright (C) 2000 by ASME

NUMERICAL METHOD
The computations are performed with the help of a code FlowER solver of compressible viscous flows through multi-stage turbomachinery developed by Yershov & Rusanov 1996ab, see also Rusanov & Yershov 1996, Yershov et al. 1998. The code has been tested
on a number of turbomachinery geometries and flow regimes for real
full-scale and model turbines, including tests held at the ERCOFTAC
Workshops on Turbomachinery Flow Prediction, see Yershov et al.
1997. The solver draws on the set of thin-layer Reynolds-averaged
Navier-Stokes equations for perfect gas. The effects of turbulence are
taken into account with the help of a modified algebraic model of
Baldwin-Lomax without the wall function. The governing equations
are solved numerically based on the Godunov-type upwind differencing and high resolution ENO scheme for the calculation of convective
derivatives, assuring second-order accuracy everywhere in space and
time, and third-order accuracy locally. The computational domain extends also on the radial gap above the unshrouded rotor blade tips,
allowing the calculation of tip leakage and its interaction with the
main stream. The following boundary conditions are incorporated: noslip and no heat flux at the walls; span-wise distribution of the total
pressure, total temperature and flow angles at the inlet to the stage,
and static pressure at the exit. The computations carried out in one
blade-to-blade passage of the stator and rotor converge to a steady
state, with the condition of spatial periodicity and mixing plane assumed. The assumed inlet/exit boundary conditions impose the pressure drop and let the mass flow rate be resultant. The comparative calculations are performed for the same pressure drop across the stage.
TESTED LP TURBINE STAGE
The tested LP stage of a steam turbine has the aspect ratio span/diameter changing throughout between 0.25-0.35, and operates
under conditions: pressure drop of about 0.2-0.5, inlet temperature 340-350K, volumetric flow rate - 400-1200 kg/s, exit dryness fraction
x=0.92-0.98. The calculations are carried out for perfect steam assuming the specific heat ratio =1.08-1.09 and gas constant R=440450, on an H-type grid of 600 000 cells refined at the endwalls, blade
walls and at the leading and trailing edges of the blades, see Fig. 1.
Computations of the original construction serve as a point of departure
for the present investigations of the effect of compound lean at the
root. Due to the fact that the last LP steam turbine stages operate in a
wide range of flow rates, the analysis should be carried out not only
for optimum operating conditions but also for low and large flow
rates. It is quite straightforward that the flow patterns exhibit significant changes with the flow regime, and also with the blade span as a
result of large gradients of pressure, velocity and swirl angle spanwise, the most significant being those gradients downstream of the
stator in the axial gap between the stator and rotor as they determine
the inlet flow conditions for the moving blade row. The above mentioned effects can be described by turbine engineers as a gradient of
stage reaction that is of significant consequence for the turbine characteristics. The effect of reaction on the flow near the root sections,
which the investigated compound lean will be most likely to modify, is
explained in Fig. 2 showing Mach numbers in the stator and rotor for
three volumetric flow rates - 420, 796 (near nominal) and 1090 m3/s.
The shape of contours discloses the location of subsonic and supersonic flow regions, and the configuration of shock waves. For the
considered flow rates and gradient of reaction supersonic velocities are
certain to occur in the stator at the root. For a low flow rate of 420

Fig. 1. Computational grid for the stator and rotor in meridional view
(top) and in blade-to-blade view 6% blade span from root (bottom).
Note that meridional and blade-to-blade grids are in different scales.
m3/s, some distance upstream of the trailing edge near the suction surface of the stator, the flow should be accompanied by week shocks
although this is not exactly clear from the pictures due to relatively
coarse resolution of the grid. With the increasing flow rates (796 and
also 1090 m3/s), the flow in the stator exhibits the occurrence of an
oblique shock wave near the trailing edge from the pressure surface
with some reflection from the suction surface, and another oblique
shock from the suction surface separating the main stream from the
wake. For 420 m3/s, the rotor is free from supersonic velocities. The
flow at the root of the rotor has an extended zone of low velocities
along the entire suction surface, with a large zone of reverse flow and
wide wake as a result of low reaction of the stage at the root. With the
increasing flow rates, interesting phenomena are observed in the rotor
at the root where an inlet shock wave appears at the suction surface.
Interaction of the inlet shock with the boundary layer induces separation downstream. Supersonic velocities are not exceeded near the trailing edge at the root for the nominal flow rate 796 m3/s, but this is not
true for the large flow rate 1090 m3/s where oblique trailing edge
shocks with reflections occur.
The observed features have a major influence on the distribution
of losses in the stage, see Fig. 3 showing the span-wise distribution of
reaction, kinetic energy losses in the stator, rotor and stage (defined in
Appendix) for three volumetric flow rates. The tendency that high velocities are observed in the stator at the root translates onto increased

Copyright (C) 2000 by ASME

Fig. 2. Mach number contours in the stator (top) and rotor (bottom) at 6% blade span from the root in the original stage
for volumetric flow rates of 420 (left), 796 (centre) and 1090 (right) m3/s.

Fig. 3. Stage reaction (1), kinetic energy losses in the stator (2) and referred to the enthalpy drop across the stage (3),
kinetic energy losses in the rotor (4), kinetic energy losses in the stage without (5) and with (6) the exit velocity
in the original stage for volumetric flow rates of 520 (left), 796 (centre) and 1090 (right) m3/s.
and span-wise distribution of losses will be compared with those for
the stage modified by compound lean at the root.
There are also interesting effects at the opposite endwall, including
the effect of tip leakage, but they are of no relevance to our investigations since compound lean at the root is most likely to modify the flow
in the root sections. By investigating the effect of compound lean at

flow losses there (endwall and blade wall boundary layer losses and
supersonic flow losses), increasing there with the increasing flow rate.
For low flow rates in the rotor, increased losses at the root are due to
underloading of rotor blades and extended separations. For the nominal and large flow rates, inlet shocks and shock-induced separation
increase flow losses at the root. The presented flow pictures at the root

Copyright (C) 2000 by ASME

the root on performance of the stage we are aiming at improving the

flow through the root passage of the overloaded stator and underloaded rotor.

COMPOUND LEAN AT ROOT - EFFECTS

The results of modification of flow patterns depend greatly on the
assumed form of compound lean. In this chapter, span-wise distributions of pressure and velocity, flow patterns at the root in the stator
and rotor, and span-wise distribution of reaction and losses for a stage
with compound lean of stator blades for the circumferential displacements of the root sections relative to mid-span equal to x=0.14l,
y=0.5l (l - blade height) are compared with the original stage. The
displacement parameters are easily explained with the help of Fig. 4.

COMPOUND LEAN AT ROOT - GEOMETRY

Compound lean at the root of LP stator blades in circumferential
view is presented in Fig. 4. It is a design with 3D blade stacking where
the root sections of the stator blades are non-linearly displaced in the
circumferential direction relative to mid-span sections so that both the
leading and trailing edges of the blades assume curvature span-wise in
circumferential view. It is important to note that compound lean does
not alter the blade in meridional view. The curvature of the leading
and trailing edges in circumferential view is assumed here in a parabolic form. It is also assumed that the pressure surface of the stator
blade meets the root endwall at an acute angle. The reader views Fig. 4
as if looking downstream, and the moving blades rotate clockwise.

Span-wise redistribution of pitch-averaged pressure and

velocity in the stator and rotor
With the assumed direction of compound lean at the root, the pressure surface of the 3D stator blade looks down to the root endwall.
Through the stator, there is a pressure rise at the root sections, compared to the stage with original stator blades; the velocity with its axial
component tend to decrease at the root, see Fig. 5a (non-dimensional
pressure is pressure referred to the critical density and critical velocity
in second power). These changes in the velocity act to reduce boundary layer losses at the root endwall and at the blades (especially on the
suction surface) at the root sections as the boundary layer losses can
be assumed proportional to the integral from the freestream velocity in
third power over the wall area, Denton 1993. Lower velocities at the
root of the stator blade also mean less risk of appearance of shock
waves, or weaker shock waves. This means less entropy creation due
to shocks, which varies with the cube of Ma2-1 (Ma - Mach number),
and further reduction of supersonic flow losses in the stator.
Compound lean in the stator has an influence on flow in the rotor.
Through the stator, the streamlines assume curvature and stray away
from the root endwall at the inlet, turning towards the root in the rotor.
As a result of changing streamwise curvature, the velocity and its axial
component are increased, and towards the exit, the pressure at the root
is observed to decrease below that of the original stage without comp-

Fig. 4. Compound lean at root

Fig. 5a. Span-wise circumferentially averaged distribution of static pressure (left), velocity (centre) and axial velocity (right) in the stator
95% axial chord downstream of the leading edge - 1 - without stator blade compound lean, 2 - with stator blade compound lean

Fig. 5b. Span-wise circumferentially averaged distribution of static pressure (left), velocity (centre) in the rotor 70 and 95% axial chord downstream
of the leading edge and axial velocity (right) 70 % axial chord downstream of the leading edge - 1- without, 2- with stator blade compound lean

Copyright (C) 2000 by ASME

Fig. 6. Mach number contours in the stator (top) and rotor (bottom) at 6% blade span from the root with stator blade compound lean
for volumetric flow rates of 428 (left), 809 (centre) and 1109 (right) m3/s
ound lean. More mass is passed through root sections in the rotor, Fig.
5b. Fortunately, this should reload the underloaded root sections of the
rotor for low flow rates. Less fortunately, it will also give more load
for large flow rates. Note that Figs. 5a and 5b are plotted for nominal
operating conditions.
Changes in flowfields
Mach number contours in the stator and rotor of the stage with
compound leaned stator blades for three volumetric flow rates - 428,
809 and 1109 m3/s are presented in Fig. 6. These pictures should be
compared with the respective pictures in Fig. 2. Slight corrections in
the volumetric flow rates follow from the fact that the comparative
computations were carried out for the same pressure drops, and it was
found that the stage with compound leaned blades is more open for
flow. However, the discrepancies in volumetric flow rate do not exceed 2% and this does not make the comparison any worse.
In the stage with compound lean there are lower velocities in the
stator at the root, compared to the original stage. This is clear from the
comparison of numbers in the key for the pictures in Fig. 6, compared
to Fig. 2. Shocks are less intensive. A weak oblique shock from the
pressure side of the blade appears only for large flow rates, whereas
for the original stage a strong oblique shock is present at nominal conditions. The most considerable qualitative changes of flow patterns
take place in the rotor at the root. Due to more mass pushed to the
root, the large separation zone over the suction surface of the blade for
low flow rates does not exist any more, or is reduced to a thin zone at
the rear part of the blade. For nominal flow rates, lower velocities
downstream of the stator blade, with the rotor rotational speed subtracted, do not yield supersonic velocities at the leading edge of the
rotor blades and the medium flows smoothly without shock-induced
separation past the front part of the suction surface. However, unlike
for the original stage for this flow rate, the supersonic velocities are

Fig. 7. Velocity vectors in the rotor at mid-blade-to-blade distance

without stator blade compound lean (left) for a volumetric flow rate of
360 m3/s and with stator blade compound lean (right) for 366 m3/s

Fig. 8. Helicity contours at the trailing edge of the stator in the original
stage (left) and with compound lean at the root (right)

Copyright (C) 2000 by ASME

Fig. 9. Stage reaction (1), kinetic energy losses in the stator (2) and referred to the enthalpy drop across the stage (3),
kinetic energy losses in the rotor (4), kinetic energy losses in the stage without (5) and with (6) the exit velocity
for the stage with stator blade compound lean (bottom) for volumetric flow rates of 527 (left), 809 (centre) and 1109 (right) m3/s.
As the blades are long, the formation of secondary flows has a minor effect on flow losses. An additional radial pressure gradient between the root endwall and mid-span sections in the stage with compound lean causes increased convection of endwall boundary-layer
fluid towards mid-span sections and there is some redistribution in the
shape of local maxima of losses due to secondary flows. This phenomenon important in HP stages with compound leaned blades can also be
noticed here, see contours of helicity (a scalar product of velocity and
vorticity) in Fig. 8, but is of little significance for the LP blades.
Fig. 10. Span-wise circumferentially averaged distribution
of the relative swirl angle at the inlet to the rotor (left) and absolute
swirl angle at the exit from the rotor (right) - 1 - without stator
blade compound lean, 2 - with stator blade compound lean

Changes in reaction and distribution of losses

Similar to Fig. 3 for the original stage, Fig. 9 shows the span-wise
distribution of reaction, kinetic energy losses in the stator, rotor and
stage for the stage with compound lean for three volumetric flow rates,
corresponding to those for which Fig. 3 were plotted. The most spectacular changes are observed near the root. The reaction at the root
tends to rise in a spectacular way by 20-30% depending on the flow
rate, leaving a minimum of the distribution away from the root endwall for a considerably large compound lean displacements as in our
case. The compound lean at the root has also some favourable effect
on the reaction at the tip, bringing it down by one or two percentage
points depending on the flow rate, but this does not seem to affect rotor and stage loss peaks due to tip leakage. Stator losses are conspicuously decreased at the root due to decreased velocities there. The same
for the rotor loss for low flow rates, also for nominal conditions, but
not for large flow rates. The summary stage losses without the exit velocity (stator plus rotor losses) are likely to be decreased for all tested
flow rates. However, the stage losses with the exit velocity do not
compare that well. First, we have to remember about a slight increase
in the volumetric flow rates for the stage with compound lean, compared to the original stage, in computations carried out for the same
pressure drops. Second, changes in the exit velocity are also due to the
changing exit swirl angle at the root which is explained in Fig.

exceeded at the trailing edge with oblique shock waves and separation
over the rear part of the blade. There are no inlet shocks even for large
flow rates, however, the situation at the trailing edge turns more dramatic. High supersonic velocities give rise to a configuration of intensive shocks interacting with the boundary layers and wakes. A separation from the root downstream of the trailing edge can be expected.
The relative exit angle is also clearly changed there.
An interesting comparison of flow patterns is presented in Fig. 7.
There are velocity vectors in the rotor at mid-blade-to-blade distance
in the original stage and in the stage with stator blade compound lean
for very low volumetric flow rates, 360 and 366 m3/s respectively, below the typical operating conditions. However, these conditions can
also occur and in the original design are pronounced in a massive
separation over some considerable part of the flow channel span-wise
when the fluid particles at the root lose their momentum and fail to
complete their way to the exit. It is clear that stator blade compound
lean at the root is a nice design to correct this failure (warning! velocity vectors in the pictures are not arrows but dots with stretches showing the direction of the velocity).

Copyright (C) 2000 by ASME

10

STATOR+ROTOR LOSS

25

STATOR LOSS

REACTION AT ROOT [%]

50

-25
600

800

1000

1200

400

VOLUMETRI C FLOW RATE

400

600

800

1000

400

1200

600

70

800

1000

1200

VOLUMETRI C FLOW RATE

VOLUM ETRI C FLOW RATE

27
11

60

50

STAGE LOSS

ROTOR LOSS

REACTION AT TIP [%]

26

10

25

24

23

22

21

40

20

400

500

600

700

800

900

1000

VOLUM ETRI C FLOW RATE

1100

1200

400

600

800

1000

VOLUMETRI C FLOW RATE

1200

400

600

800

1000

1200

VOLUMETRI C FLOW RATE

Fig. 11. Change of stage reaction at the root (top left) and tip (bottom left), stator loss (top centre) and rotor loss (bottom centre), stage loss without
the exit velocity (top right) and stage loss with the exit velocity (bottom right) as a function of volumetric flow rate due to the effect of stator blade
compound lean at the root for different lean displacements; ! - x=0.0, y=0.0 - original construction; " - x=0.08, y=0.4; 5- x=0.14, y=0.5;
#- x=0.2, y=0.6. Minimum division on the axis of losses is 1.0%.
10 showing the comparison of span-wise circumferentially averaged
distribution of relative swirl angle at the inlet to the rotor and absolute
swirl angle at the exit from the rotor without and with stator blade
compound lean. For the presented stage, stator blade compound lean
reduces the swirl angle downstream of the stator blades at the root
which can be good for the design reduction of the rotor blade twist.
On the other hand, the changing streamline curvature pushes more
mass to the root in the rotor and there is an increase in the negative
swirl angle downstream of the rotor blade which will increase the exit
velocity loss (the swirl angle is measured to the axial direction).
The compound lean at the root modifies distributions mostly at the
root. There are hardly any changes in distributions of stator and rotor
losses in sections above the mid-span. The distribution of rotor and
stage losses presented in Fig. 9 refers to the section located 50% of the
axial chord downstream of the rotor trailing edge. There will certainly
be more loss further downstream as the mixing processes are not yet
accomplished at the assumed test section.

not affect rotor and stage loss peaks due to tip leakage for the compound lean displacements x=0.14l, y=0.5l, but this does not mean
that favourable effects of reduction of tip leakage can not appear for
larger displacements.
Stator losses are seen to decrease with the increasing compound
lean displacements which is especially clear for large flow rates. This
tendency is also evident in the rotor for low flow rates. However,
above the nominal conditions the tendency is broken and for large
flow rates, stator blade compound lean can bring efficiency losses in
the rotor. The summary stage losses without the exit velocity (stator
plus rotor losses) seem to be evenly decreasing with the increasing
compound lean displacements over the tested range of volumetric flow
rates, but not the stage losses with the exit velocity, as the exit velocity
losses exhibit a tendency to increase for the stage with compound
leaned blades, compared to the original stage, see previous chapter.
Stage losses with the exit velocity depend greatly on the flow regime. A reduction by 1% is observed in the region of low load and
even higher efficiency gains should be expected for yet lower loads.
Computations were also made for those lower loads below the usual
operating conditions, however the obtained numbers of efficiency
gains will not be plotted here. This follows from the fact that the computations exhibit large separation zones at the exit from the rotor as in
Fig. 7, and exact computations of the flow field would require
exit/inlet conditions at the otherwise exit section of the rotor. If not so,
also the determination of flow losses can be charged with uncertainty.
More certain, the design is less favourable for high flow rates, due to
the increased rotor and exit velocity losses. There are practically no
efficiency gains for volumetric flow rates between 900-1100, and efficiency losses are likely for larger flow rates. It is important to note that
the presented characteristics account for the effect of changing flow
rates with changing compound lean displacements for the same pressure drop, a fact not considered in earlier comparisons.

COMPOUND LEAN - CHARACTERISTICS

In this chapter, investigations will be conducted for several configurations of stator blade compound lean for different displacements
parameters: 1 - x=0.0, y=0.0, 2 - x=0.08l, y= 0.4l, 3 - x=0.14l,
y=0.5l, 4 - x=0.20l, y=0.6l, where l is the blade height. Fig. 11
shows the change of stage reaction at the root and tip, stator and rotor
loss, stage loss without and with the exit velocity due to the effect of
stator blade compound lean at the root as a function of volumetric
flow rates, for different lean displacements as listed above. As already
mentioned, the reaction at the root rises considerably with the increasing compound lean displacements. It seems that modifications at the
root can also have an effect on the reaction at the tip which is slightly
decreased with the increasing lean displacements. As shown, this did

Copyright (C) 2000 by ASME

Process Simulation, Espoo, Finland, June 13-16.

Yershov S.V., Rusanov A.V., 1996b, The application package
FlowER for calculation of 3D viscous flows through multi-stage turbomachinery, Certificate of Ukrainian state agency of copyright and
related rights, Kiev, Ukraine, February 19.
Yershov S.V., Rusanov A.V., Gardzilewicz A., Badur J., Lampart
P., 1997, Calculations of Test Case 3 - Durham low speed turbine cascade, Calculations of Test Case 9 - Highly loaded transonic linear turbine guide vane cascade, Proc. V ERCOFTAC Seminar and Workshop
on 3D Turbomachinery Flow Prediction, Courchevel, France, January
6-9.
Yershov S.V., Rusanov A.V., Gardzilewicz A., Lampart P.,
wirydczuk J., 1998, Numerical simulation of viscous compressible
flows in axial turbomachinery, TASK Q., Vol. 2, No. 2.

CONCLUSIONS
The effect of stator blade compound lean at the root on flow characteristics of an LP turbine stage has been investigated numerically
with the help of a viscous compressible 3D solver FlowER. It has been
found that compound lean induces additional blade force and streamwise curvature, as well as redistribution of flow parameters in the
stage. The design can improve flow patterns at the root sections of the
overloaded stator and underloaded moving blade row, and reduce flow
losses. Efficiency gains depend on the flow regime, with the highest
gains in the region of low load. However, this design can be unfavourable for high mass flow rates. The presented numbers for efficiency
gains are assumed valid for the considered stage and can change with
stage geometry.
ACKNOWLEDGEMENTS
The author would like to thank Prof. Andrzej Gardzilewicz from
Institute of Fluid Flow Machinery, Gdask, Poland, as well as Prof.
Sergey Yershov and Dr. Andrey Rusanov from Institute of Mechanical
Engineering Problems, Kharkov, Ukraine, for their encouragement to
carry out investigations, helpful comments and discussions.

APPENDIX
In this chapter the flow parameters used in the paper, including the
reaction, kinetic energy losses in the stator, rotor and stage, are defined. Fig. A1, which illustrates the process of expansion in a turbine
stage in the form of an enthalpy-entropy diagram, will help to explain
the definitions gathered in Tab. A1.

REFERENCES
Denton J.D., 1993, Loss mechanisms in turbomachines, ASME J.
Turbomachinery, Vol. 115.
Denton J.D., Xu L., 1999, The exploitation of 3D flow in turbomachinery design, VKI LS 1999-02.
Harrison S., 1992, The influence of blade lean on turbine losses,
ASME J. Turbomachinery, Vol. 114.
Lampart P., Gardzilewicz A., 1999, Numerical study of 3D blading in HP impulse turbines, Presented at SYMKOM'99, Arturwekd, Poland, October 5-8, Published in Cieplne Maszyny
Przepywowe, Vol. 115.
Lampart P., Gardzilewicz A., Rusanov A., Yershov S., 1999a, The
effect of stator blade compound lean and twist on flow characteristics
of a turbine stage - numerical study based on 3D NS simulations, Presented at 2nd Symp. on Comp. Technologies for Fluid/Thermal/
Chemical Systems with Industrial Applications, ASME PVP Division
Conf., Boston, USA, August 1-5, Published in ASME PVP - Vol. 397
No. 2.
Lampart P., Gardzilewicz A., Rusanov A., Yershov S., 1999b, Radial lean and compound lean of stator blades as means of improving
flow characteristics of HP turbine stages, Proc. Fourth Int. Symp. on
Exp. & Comp. Aerothermodynamics of Internal Flows, Dresden, Germany, August 31-September 2,
Rusanov A.V., Yershov S.V., 1996, The new implicit ENO
method for 3D viscous multi stage flow calculations, Proc. 3rd ECCOMAS Computational Fluid Dynamics Conf., Paris, France, September 9-13.
Singh G., Walker P.J., Haller B.R, 1995, Development of threedimensional stage viscous time marching method for optimisation of
short height stages, Proc. Europ. Conf. on Turbomachinery, Fluid
Dynamics and Thermodynamic Aspects, Erlangen, March 1-3.
Wang Z., 1999, Three-dimensional theory and design method of
bowed-twisted blade and its application to turbomachines, VKI LS
1999-02.
Yershov S.V., Rusanov A.V., 1996a, The high resolution method
of Godunovs type for 3D viscous flow calculations, Proc. 3 Colloq.

Reaction

= ( h1 h2 s ) / ( h0T h2 s ' )

Stator loss

1 = ( h1 h1s ) / ( h0T h1s )

Stator loss referred to the enthalpy

drop across the stage
Rotor loss

1 = ( h1 h1s ) / ( h0T h2 s ' )

2 = ( h2 h2 s ) / ( h1T h2 s )

Stage loss without the exit velocity

12 = ( h2 h2 s ' ) / ( h0T h2 s' )

Stage loss with the exit velocity

12 c = ( h2 T h2 s ' ) / ( h0 T h2 s ' )

Tab. A1. Reaction, kinetic energy losses in the stator, rotor and stage.

Symbols:
c - absolute velocity,
h - enthalpy,
p - pressure,
s - entropy,
w - relative velocity,
Subscripts:
0 - inlet
1 - behind the stator
2 - behind the stage
s, s - isentropic
T - total

Fig. A1. Enthalpy-entropy diagram for a turbine stage.

Copyright (C) 2000 by ASME