Lecture 2 Rothalpy
Lecture 2 Rothalpy
Lecture 2 Rothalpy
∆hT = Q − W x = 0 = Cp∆To Total pressure rises through a compressor rotor and falls
through a turbine.
∴ To = const.
For a perfect gas this means that the total temperature So in the rotor of a turbomachine stagnation conditions,
is also constant. temperature and pressure are not conserved. So what
properties of the fluid are conserved in a rotor?
If the flow is also reversible (lossless) the second law The Euler Work Equation Applies:
tells us:
δP ∆hT = ∆UCθ
Toδs = δho − o ; δs = 0, δho = 0
ρo hT 2 − hT 1 = U 2Cθ 2 − U1Cθ 1
∴ δPo = 0 i.e. Po = const. i.e. hT − UCθ = const. (1)
γ γ
PT (γ − 1) 2 γ −1 PTrel (γ − 1) γ −1
= 1 + M = 1 + M rel
2
P 2 P 2
Slide Number 6 Dr. M.G. Rose
“Loss” in Turbomachinery
T P
Traditionally the term loss has been taken as loss of
s2 − s1 = Cp ln 2 − R ln 2
either efficiency or total pressure ∆PT. However, T1 P1
from a thermodynamic stand point we should not
consider lost total pressure but lost work. Lost
work is defined as a reduction of the steady flow Efficiency Definitions
availability function (B): There are a number of different definitions of
B = h − Tatm s efficiency used in turbomachinery. The most
common are defined here.
In this expression Tatm is a datum environmental Efficiencies compare the actual changes in the
temperature. If the machines are essentially
fluid between inlet and exit with the theoretical
adiabatic then lost work is due to entropy rise.
best ‘reversible’ change. Also at exit either
So to address the topic of loss in turbomachinery total or static conditions can be used. This
we need to study the mechanisms which cause gives either ‘Total to Total’ or ‘Total to Static’
increases in entropy in the flowing fluid. efficiency. The Difference is the exit kinetic
energy of the gas.
Loss
Loss==Entropy
EntropyRise
Rise Another efficiency definition used is the
Polytropic efficiency defined on the next foil.
An entropy rise always leads to a reduction of
efficiency.
∆Po ∆Po 2
Y= or ηrow
Cact
= 2
P D Cideal
∆Po is the loss of total pressure across the blade row. P is
the inlet total pressure, D is the inlet dynamic head for a The ideal exit velocity is based on the inlet total
compressor row and the exit dynamic head for a turbine pressure and exit static pressure
row. These loss coefficients only capture the entropy rise
γ −1
due to reductions in total pressure. Typical values for these
2 Po1 γ
2
= −
γ − 1 P2
loss coefficients for a 2D cascade flow 2% to 4% M ideal 1
Compressor Turbine
P −P P −P 2
= M ideal
2
(γRTideal )
Y = o1 o 2 Y = o1 o 2 Cideal
Po1 − P1 Po 2 − P2
T
C 2
ideal = M T γR ideal
2
ideal o1
Blade Row Efficiency To1
For Turbines a blade row efficiency is often
1
quoted. This compares the actual exit kinetic 2
Cideal = M ideal γRTo1
2
1 + γ −
M ideal
energy with the lossless exit kinetic energy at the 1 2
same exit static pressure. 2
ke
wa
Friction as a Source of Loss τw = shear stress on wall due to fluid
The aerofoils and endwalls of a turbomachine are δ = displacement thickness of boundary layer
immersed in fluids which are viscous. As the fluid
moves past the surfaces shear stresses are created which ∂u
τ w = µ
∂y wall
oppose the motion of the fluid. This occurs in the
boundary layers which form on all surfaces. At the
bottom of the boundary layers the fluid is assumed to be Where µ is the laminar viscosity and y is the
fixed and stationary relative to the solid surface. coordinate normal to the surface.
Slide Number 11 Dr. M.G. Rose
Review of Boundary Layers
∂P ∂U
• Two Dimensional flow with friction For =0 ⇒ =0
• Zero slip at the wall ∂x ∂x
u τ ∂θ τo
∫
L
y ∴ o2 = Integrate θ = ∂x
y ρU ∂x 0
ρU 2
x
∫
L
ρU 2θ = 0
τ o ∂x ≈ drag
u
∫ 0 1 − U ∂y : Displacement Thickness
∞ In practice there is also drag due to the pressure
δ* = distribution; tailing edge drag in particular. Drag is
not a very helpful concept in turbomachinery, we
∞ u u
θ = ∫ 0 1 − ∂y : Momentum Thickness
must consider entropy rise.
U U
Simple Flat Plate Correlations
∂u
τ o = µ : Wall Shear Stress Laminar
∂y y = 0
θ
Momentum Integral Equation = 0.664 Re−x0.5
x
∂θ δ * θ ∂U τo Turbulent
+ 2 + = θ
= 0.036 Re−x0.2
∂x θ U ∂x ρU 2 x
Slide Number 12 Dr. M.G. Rose
Example Approximate Mixing Loss For a Turbine
Blade Row: Control Volume
Nomenclature
θ = Total momentum thickness, boundary layers at trailing edge Scosα2
1
δ* = Total displacement thickness, boundary layers at trailing edge 2
ρ = Density
V = Velocity
S = Pitch
Thin Trailing Edges
α = Flow angle to axial
P = Static Pressure
V2
Po = Total pressure
∆ = Denotes ‘change in’
ṁY = Mixing Loss Coefficient
= Mass flow rate
Continuity
∆Po
V12 2 δ + θ
*
( )
( )
m˙ = ρV1 S cosα2 − δ * = ρV2 S cosα2 1 ρV 2
2 2
= 1 − 2 1 −
V2 S cosα2
Momentum
From the continuity equation: if displacement
2δ * 2 δ + θ
*
( )
2 2 (
Po2 − Po1 = ∆Po = P1 + 1 ρV12 − P2 + 1 ρV22 ) Y = 1 − 1 + 1 −
S cosα2 S cosα2
Following same approximation
Substitute for P 1-P2
δ *
+θ ∆Po2θ
∆Po = 1 2 ρV − 1 2 ρV1 +
2 2
ρV12 Y= ≈
S cosα2 1 ρV22 S cosα2
2
2
Slide Number 14 Dr. M.G. Rose
For this example
The mixing loss is principally dependent on the momentum
thickness
When operated at the design point turbomachinery blading should have zero incidence, the flow arrives at the
aerofoils, stators and rotors traveling in the right direction. At all other operating conditions the blading experiences
incidence. Positive incidence always indicates more turning, negative less. The two graphs show typical response to
incidence. Boundary layer separation is often the cause of increased losses.
+ive +ive
Y = ∆Po/(ρV12/2)
Y = ∆Po/(ρV22/2)
Compressor Profile in Turbine Profile in
Cascade Effect of Cascade Effect of
Incidence Incidence
0.02 0.02
≈ +45°
α1 α1
≈-20° Zero ≈ -45° Zero
Incidence +5° to +10° Incidence
≈ 1.0 ≈ 1.25
Inlet Mach No. M1 M2
Exit Mach No.
Shocks form around trailing edge and interact with
Shocks form around leading edge ‘bow wave’ adjascent suction side. Once the axial exit mach
and interact with suction side boundary layer number reaches sonic there can be no further
where they may cause separation of the boundary changes of lift. The turbine is in ‘limiting output’
layer. condition.
Wakes
Shock Possible separation
Waves Separation
Throat M ≈ 1.0
Shock Waves
Mach number
flows and unsteadiness all tend to suppress laminar flow.
Transition is one of the hardest characteristics of the flow in
turbomachinery to predict. Both shear stresses and heat transfer
are much greater in the turbulent regions. One can also diffuse
much more without separation in a turbulent flow.
The leading edges of the aerofoils have new laminar boundary Cax
layers. Transition normally occurs somewhere along each Stagnation laminar
surface. If one can maintain a laminar flow field higher Laminar
efficiencies can be achieved. Separation
Transition Factors
Laminar
Laminar Separation (LE or SS) Curvature Effects Separation
Diffusion Film Cooling Effects
Reattachment Wake Interaction Turbulent
Reattachment
Natural Transition Reynolds Number
In the real unsteady flow found inside a machine the
Relaminarisation Becalmed Regions
wakes from upstream cause unsteady transition to
Roughness occur. This can supress the laminar separation and
introduce low loss becalmed regions.
Most boundary layers inside turbomachines are The losses directly due to the shocks are quite
turbulent. A simple expression exists for the modest e.g. for M1 = 1.5 a normal shock:
entropy rise in a turbulent boundary layer. Po 2
= 0.93
Loss ∝ ∫ U 3 ∂A
Po1
This is a very strong shock with a static
This is the surface integral of the cube of the pressure ratio of about 2.5. However, the losses
velocity outside the boundary layer†. This caused by shocks can be quite large. Shocks
shows that regions of high velocity generate can cause boundary layers to separate and
much more loss than regions of slightly lower thereby increase mixing losses. If shocks
velocity. A factor of 2 in velocity gives nearly oscillate due to unsteady interaction additional
an order of magnitude on loss! For this reason losses result. Also the shock interaction
most of the loss on a 2D aerofoil is on the between stators and rotors can result in big
suction side. About 80% of the loss is found on variations of work and massflow, which
the suction side. subsequently mix out causing extra loss.
Cumpsty (1989)
Passage Vortex
Gregory-Smith 2002
Slide Number 22 Dr. M.G. Rose
Non-Dimensionalised Total Pressure Loss Experimental Traverse Results 28%
Downstream of Turbine Cascade
Wake Gregory-Smith 2002
“Loss Core”
“Passage Vortex”
Takeishi 1988
Flow Shroudless
Loss of Efficiency %
Stator
2.0
S
δ = k *θ
C
Where k*= constant (0.2 typical), θ camber angle
(difference between metal angles), S pitch, C true
chord. Typical value of δ in compressors is 8°.
Slide Number 28 Dr. M.G. Rose