FDP on Fluid Mechanics and Machinery
Axial Flow Turbines
K M Parammasivam
Professor
Dept of Aerospace Engineering
MIT Anna University
Axial Flow Turbines
• Working fluid is accelerated
by the stator and decelerated
by the rotor
• Boundary layer growth and
separation does not limit
stage loading as in axial
compressor
• Expansion occurs in stator and in relative frame of rotor
Elementary Theory
Perfect
• Energy equation for control V32 V12
gas
q w h3 h1 h03 h01 c p T03 T01
volumes (again): 0
2 2
• Adiabatic expansion process h03 h01
(work extracted from system -
sign convention for added
work = +w)
– Rotor => -w = cp(T03-T02) <=>
w = cp(T02-T03)
– Stator => 0 = cp(T02-T01)
=> T02= T01
How is the temperature drop related to the blade angles ?
• We study change of angular
momentum at mid of blade
(as approximation)
Governing Equations and Assumptions
• Relative and absolute refererence
frames are related by: V C U
• We only study designs where:
– Ca2=Ca3
– C1=C3
• You should know how to extend the equations!!!
• We repeat the derivation of theoretical work used
for radial and axial compressors:
Rate of change of angular momentum
Cw3 r3 Cw 2 r2 Theoretica l torque
Theoretica l work Cw3 r3 Cw 2 r2 Cw3U 3 Cw 2U 2
Assume changeof
direction for relative
velocity
Flow at constant radius U Cw3 Cw2 U Cw3 Cw 2
Principle of Angular Momentum
Stage work output w:
w U Cw2 Cw3
U Ca 2 tan 2 Ca 3 tan 3
Ca constant:
w U Ca 2 tan 2 Ca 3 tan 3
UCa tan 2 tan 3
Energy equation
Combine derived equations =>
w UCa tan 2 tan 3
Energy equation:
w c p T0,stage
We have a relation between temperature drop
and blade angles!!! :
c p T0,stage UCa tan 2 tan 3 (7.3)
Exercise: derive the correct expression when 3 is small enough to allow
3 to be pointing in the direction of rotation.
Dimensionless parameters
Blade loading coefficient, temperature drop
coefficient:
c p T0, stage
1 2
Equation 7.3
2Ca
tan 2 tan 3 (7.6)
U U
2
T2 T3
Degree of reaction:
T1 T3
T2 T3
Exercise: show that this expression is equal to =>
when C3= C1 T01 T03
can be related to the blade angles!
C3 = C1 =>
c p Tstage c p T0,stage UCtan 2 tan 3
Relative to the rotor the flow does no work (in the relative frame
the blade is fixed). Thus T0,relative is constant =>
c p Trotor V3 V2 Ca tan 3 tan 2
1 2 2 1 2 2 2
2 2
Exercise: Verify this by using the definition V2
T0,relative T
of the relative total temperature: 2c p
can be related to the blade angles!
Plugging in results in definition of =>
T2 T3 Ca
tan 3 tan 2 (7.7)
T1 T3 2U
The parameter quantifies relative amount of ”expansion” in
rotor. Thus, equation 7.7 relates blade angles to the relative
amount of expansion. Aircraft turbine designs are typically
50% degree of reaction designs.
Dimensionless parameters
Finally, the flow coefficient:
Ca
U
Current aircraft practice (according
to C.R.S):
0.8 1.0
3.0 5.0
0.5
Aircraft practice => relatively high
values on flow and stage loading
coefficients limit efficiencies
Dimensionless parameters
Using the flow coefficient in 7.6 and 7.7 we obtain:
2 tan 2 tan 3 (7.8)
tan 3 tan 2 (7.9)
2
The above equations and 7.1 can be used to obtain the gas
and blade angles as a function of the dimensionless parameters
1 1 1
tan 2 2 tan 2 tan 2
2 2
1 1 1
tan 3 2 tan 3 tan 3
2 2
Simple Problems
• Exercise: show that the velocity triangles become symmetric for
= 0.5. Hint combine 7.1 and 7.9
• Exercise: use the “current aircraft practice” rules to derive bounds
for what would be considered conventional aircraft turbine designs.
What will be the range for 3? Assume = 0.5.
Turbine loss coefficients:
Nozzle (stator) loss coefficients:
T2 T2
N 2
C2
2c p
p01 p02
YN
p02 p2
Nozzle (rotor) loss coefficients:
T3 T3
R
V22
2c p
p02,rel p03,rel
YR
p03,rel p3
3D design - vortex theory
• U varies with radius
• Cw velocity component at
stator exit => static pressure
increases with radius =>
higher C2 velocity at root
• Twist blades to take changing
gas angles into account
– Vortex blading
3D optimized blading
(design beyond free
vortex design)
3D design in steam turbines
• Keep blade angles from
root to tip (unless rt/rr high)
• Cut cost
• Rankine cycle relatively
insensitive to component
losses
Choice of blade profile, pitch and chord
• We want to find a blade that will minimize loss and perform the required
deflection
• Losses are frequently separated in terms:
Measured Groupedinto one term
in cascade secondaryloss s
Total Profile Tip clearance Annulus Secondary
Loss flow loss
Choice of blade profile, pitch and chord
• As for compressors - profile families are used for thickness distributions.
For instance:
– T6, C7 (British types)
Choice of blade profile, pitch and chord
• Velocity triangles determine gas
angles not blade angles.
– arccos(o/s) should approximate
outflow air angle:
• Cascade testing shows a rather
large range of incidence angles for
which both secondary and profile
losses are relatively insensitive
Choice of blade profile, pitch and chord
• Selection of pitch chord:
– Blade loss must be minimized (the greater the
required deflection the smaller is the optimum
s/c - with respect to λProfile loss)
– Aspect ratio h/c. Not critical. Too low value =>
secondary flow and tip clearence effects in large
proportion. Too high => vibration problems
likely. 3-4 typical. h/c < 1 too low.
– Effect on root fixing
• Pitch must not be too small to allow safe fixing
to turbine disc rim
