Engines · Chapter 16
Compressors Centrifugal and axial compressors: how they raise pressure, why they stall and surge, and what designers do about it.
pressure rises a little at every stage
rotor
stator
the annulus converges
© 2026 eatpl.in. For the personal study of eatpl.in students only. Do not copy or distribute.
e-books by eatpl.in · © 2026 · Do not copy or distribute
Welcome to compressors. Every gas turbine depends on its compressor. Before fuel can be burnt usefully, the air has to be squeezed, often to thirty or forty times its inlet pressure, and the compressor uses more than half of the power the turbines produce to do it. In the engine on the right the air is compressed by the fan, then by an intermediate compressor, and finally by a high pressure compressor, each on its own shaft. In this lesson you will learn how centrifugal and axial compressors work, what happens to pressure, temperature and velocity in every row of blades, why compressor blades stall and how a stall grows into a surge, and the devices designers use to prevent it.
combustion
turbine
THE COMPRESSOR: this lesson
pressure ratio up to 35 : 1 or more
air leaves at up to about 600 °C
1 · Two types Pressure ratio, the centrifugal compressor, impeller and diffuser, double entry and two stages.
2 · Axial compressors Rotor and stator stages, pressure, temperature and velocity, the annulus and blade twist.
3 · Stall and surge The velocity triangle, causes and indications, off-design running and the compressor map.
4 · Control and build VIGVs, variable stators, bleed valves, spools, clearance control, blades, fans and washing.
Chapter 16 · Compressors
2
e-books by eatpl.in · © 2026 · Do not copy or distribute
This lesson has four parts. First we define the pressure ratio and look at the older, simpler type of compressor, the centrifugal compressor, with its impeller and diffuser. Second, the axial compressor that dominates large engines today: what a stage is, exactly what happens to pressure, temperature and velocity in the rotor and the stator, and why the blades get shorter towards the back and are twisted from root to tip. Third, the problem that every compressor designer fights: the stall, and its big brother the surge. We will use the velocity triangle to see why they happen, and learn the causes, the indications and the correct pilot action. Finally, the cures, from variable vanes and bleed valves to multiple spools and clearance control, followed by how compressors are built and kept clean.
Part 1 · Two types
The job, and the pressure ratio Air must be compressed before fuel is added, burnt and expanded through the turbines The compressor is driven by a turbine through a shaft Pressure ratio = compressor outlet pressure ÷ compressor inlet pressureCompressing air does work on it, so its temperature rises with its pressure At 35 : 1 the delivery air reaches about 600 °C
Pressure ratio = Poutlet ÷ Pinlet
e.g. 14.7 psi in, 514 psi out → 35 : 1
1
10
20
30
40
0
100
200
300
400
500
600
700
pressure ratio
Delivery temperature (°C), 15 °C day
4:1 ≈ 180 °C
35:1 ≈ 610 °C
real compressor
ideal (no losses)
Squeezing air heats it: the work done appears as temperature
Chapter 16 · Compressors
3
e-books by eatpl.in · © 2026 · Do not copy or distribute
The air must be compressed before fuel is added to it in the combustion chambers, and before the hot gas is expanded through the turbines. The compressor is driven by a turbine, connected to it by a shaft. We measure how hard a compressor works by its pressure ratio: the pressure at the compressor outlet divided by the pressure at the compressor inlet. If air arrives at fourteen point seven psi and leaves at five hundred and fourteen, the ratio is thirty five to one. Notice that the ratio is not the number of stages, and it is nothing to do with the exhaust. Squeezing air also heats it, because the work done on the air has to go somewhere. The graph shows the delivery temperature on a fifteen degree day. A real compressor at thirty five to one delivers air at around six hundred degrees Celsius, and that hot air is the starting point for the combustion chamber.
Part 1 · Two types
Centrifugal and axial compressors
Centrifugal
Axial
eye
impeller
diffuser
outlet elbow, cascade vanes
• robust, simple, cheap to make
• about 4 : 1 from one stage
• small mass flow for its frontal area
rotor
stator
• many stages: 35 : 1 or more overall
• large mass flow for a small frontal area
• more delicate, costly, more prone to FOD
For the same frontal area , the axial compressor swallows far more air and reaches a far higher pressure ratio, so it gives more thrust.
Chapter 16 · Compressors
4
e-books by eatpl.in · © 2026 · Do not copy or distribute
There are two basic types of compressor. In the centrifugal compressor, on the left, the air enters the middle, is flung outwards by a spinning impeller and then slowed down in a diffuser, so the flow leaves radially. In the axial compressor, on the right, the air flows straight along the engine through many rows of blades. The centrifugal compressor is much more robust, simpler, and easier and cheaper to make, which is why early jet engines used it, and why it is still found in small engines, auxiliary power units and turboprops. Its big drawback is mass flow. Compare two compressors with the same frontal area: the axial one swallows far more air and reaches a far higher pressure ratio. Thrust depends on the mass of air flowing through the engine, so for large engines the axial compressor wins. Some engines combine the two, with axial stages followed by a centrifugal stage.
Part 1 · Centrifugal
Impeller and diffuser Air enters the eye (centre) of the impeller and is flung outwards to the tip The impeller adds energy: velocity and pressure rise in its divergent passages The diffuser : fixed divergent ducts turn velocity into pressure Pressure rise shared about 50 % impeller, 50 % diffuser One stage: about 4 : 1 . Single stage = impeller followed by diffuser
rotation
eye
Impeller
flings the air outward: V, p and T rise
Diffuser
divergent passages: V falls, p rises
Chapter 16 · Compressors
5
e-books by eatpl.in · © 2026 · Do not copy or distribute
Let us look at the centrifugal compressor in detail. A turbine spins the impeller at very high speed. Air is drawn continuously into its eye, the centre, through guide vanes, and centrifugal force sends it outward along the vanes towards the tip. The passages between the vanes diverge, and the impeller is adding energy, so the air leaves the tip with both a higher pressure and a much higher velocity. It then enters the diffuser, a ring of fixed, divergent passages that slow the air down and turn that kinetic energy into pressure energy. In practice the pressure rise is shared roughly equally, about half in the impeller and half in the diffuser. A single stage, which always means an impeller followed by a diffuser, gives a pressure ratio of about four to one. In the scene, the diffuser leads straight to the combustion chambers.
Part 1 · Centrifugal
What happens to the air in a centrifugal stage
each quantity as a share of its own maximum
intake
impeller (rotating)
diffuser (fixed)
outlet elbow
static pressure
temperature
velocity
Impeller: V ↑ p ↑ T ↑
Diffuser: V ↓ p ↑ T ↑
Pressure rise shared
about 50 : 50
Overall about 4 : 1
Air enters at the eye , is accelerated radially by the impeller and leaves the diffuser tangentially , slowed down and at higher pressure.
Chapter 16 · Compressors
6
e-books by eatpl.in · © 2026 · Do not copy or distribute
Here is what happens to the air as it passes through one centrifugal stage, worked out for a four to one compressor. Through the impeller everything goes up. The velocity rises strongly as the air is whirled out to the tip, the static pressure rises to about half of the final value, and the temperature rises too, because work is being done on the air. Then comes the diffuser. The velocity falls sharply, and that kinetic energy reappears as pressure, so the pressure carries on rising to its final value. The temperature also continues to rise slightly as the air is slowed. So remember the two patterns: in the impeller, velocity, temperature and pressure all increase; in the diffuser, velocity decreases while temperature and pressure increase. Exam questions also describe the path: in at the eye, accelerated radially, and out of the diffuser tangentially.
Part 1 · Centrifugal
Double entry, two stages and cascade vanes Double-entry impeller: vanes on both faces of one disc, still a single stage For more than about 4 : 1, put two impellers in series A third stage is impractical: tip speed and centrifugal loads too high So centrifugal engines stay below about 15 : 1 Cascade vanes in the outlet elbows turn the air through large angles with little loss and complete diffusion
Double-entry impeller (one stage)
out to the diffuser
air in
air in
Vanes on both faces: more air for the same diameter
Two stages in series: the most that is practical
stage 1
≈ 4 : 1
stage 2
≈ 4 : 1
Third stage? Tip speed and centrifugal stress too high,
so centrifugal engines stay below about 15 : 1 overall
Cascade vanes in the outlet elbows turn the air
through large angles with little loss and finish diffusion
Chapter 16 · Compressors
7
e-books by eatpl.in · © 2026 · Do not copy or distribute
There are two ways to get more out of a centrifugal compressor. The first is the double-entry impeller, which has vanes on both faces of a single disc. Air reaches it from the front and from the back, so a lot more air passes for the same diameter. Be careful in the exam: that is still a single stage, because there is only one impeller and one diffuser. The second way, to get a pressure ratio above about four to one, is to put two impellers in series, the second compressing the air already compressed by the first. A third stage has not proved practical, because the tip speeds and the centrifugal loads become too great, so engines using centrifugal compressors stay below about fifteen to one overall. At the elbows of the outlet casing you will find cascade vanes. They turn the air through large angles, around ninety degrees, with the minimum of loss, and they complete the diffusion.
Part 2 · Axial
The axial compressor stage Rows of rotor blades (on discs, turning) alternate with stator vanes (fixed to the casing) One stage = one rotor row followed by one stator row Rotor: work is added, velocity, pressure and temperature rise Stator: divergent passages turn velocity into pressure ; pressure and temperature still rise Pressure rises in both rotor and stator; velocity rises only in the rotor
One stage = a rotor row, then a stator row
rotor moves
ROTOR
STATOR
air in
out, swirl removed
Rotor
Stator
Velocity
rises ↑
falls ↓
Pressure
rises ↑
rises ↑
Temperature
rises ↑
rises ↑
Chapter 16 · Compressors
8
e-books by eatpl.in · © 2026 · Do not copy or distribute
The axial compressor does the same basic job as the centrifugal one, turning kinetic energy into pressure, but in a different way. It has many rows of rotating blades, the rotor, each followed by a row of fixed vanes, the stator, all of aerofoil section. One stage is one row of rotor blades followed by one row of stator vanes, always in that order: rotor, then stator. The spaces between the blades form divergent passages in both rows. In the rotor, which the turbine keeps turning at high speed, mechanical energy is added, so the velocity, the pressure and the temperature of the air all rise. In the stator the air is slowed down and its swirl is removed, so the velocity falls, but the pressure and the temperature continue to rise. So the rotor plays the part of the impeller and the stator the part of the diffuser. Watch the pressure graph under the engine climb stage after stage.
Part 2 · Axial
Pressure, temperature and velocity, row by row
each quantity as a share of its own maximum
R1
S1
R2
S2
R3
S3
R4
S4
stage 1
stage 2
stage 3
stage 4
static pressure
temperature
absolute velocity
axial velocity
Velocity: up in each rotor,
back down in each stator
Axial velocity held
about constant
Across the whole compressor: pressure up, temperature up, axial velocity about constant .
Chapter 16 · Compressors
9
e-books by eatpl.in · © 2026 · Do not copy or distribute
This graph follows the air through four axial stages, worked out for a ratio of one point two to one per stage. The blue pressure line rises in every rotor and in every stator: a little in each row, but always upwards. The orange temperature line does the same. The solid green line is the absolute velocity of the air, and it makes a saw-tooth: the rotor speeds the air up, the stator slows it down again, stage after stage. The dashed green line is the axial velocity, the speed along the engine, and the designer holds it roughly constant from front to back. So if an exam question asks what happens to the air through an axial compressor as a whole, the answer is: pressure increases, velocity stays about constant, and temperature increases. If it asks where the velocity increases, the answer is the rotors.
Part 2 · Axial
Small steps, many stages Each stage gives only about 1.1 to 1.2 : 1 So many stages are needed, on up to three spools High-bypass engines reach about 35 : 1 or more overall First stage: about 3 psi rise; last stage: about 80 psi Compressor outlet: up to about 600 °C ; highest pressure in the engine
0
5
10
15
20
1
10
20
30
40
number of stages
Overall pressure ratio
1.2 : 1
1.15 : 1
1.1 : 1
35 : 1
At 1.2 : 1, 35 : 1 needs about 20 stages (1.2n = 35)
First stage: 14.7 psi × 0.2 ≈ 3 psi rise
Last stage near 514 psi: ≈ 80 psi rise
Same ratio per stage, far bigger psi at the back
Chapter 16 · Compressors
10
e-books by eatpl.in · © 2026 · Do not copy or distribute
A single axial stage gives only a small pressure rise, about one point one to one point two to one. That is far less than a centrifugal stage. So to reach the pressure ratios that modern engines need, many stages are used, often split between two or three spools. Because the ratios multiply, the overall ratio grows quickly. At one point two to one per stage, about twenty stages give thirty five to one, which is typical of a large high bypass engine. Here is a neat point. The first stage, working on air at fourteen point seven psi, adds only about three psi. The last stage, at the same ratio but working on air at over five hundred psi, adds about eighty psi. The air leaves at up to about six hundred degrees. In the scene, see how the pressure peaks at station three, the compressor exit, while the temperature jumps later, in the combustor.
Part 2 · Axial
The converging annulus The air annulus is the space between the rotor drum and the casing As the air is compressed its density rises and its volume falls To keep the axial velocity constant , the annulus area must reduce Done by tapering the casing , the drum , or both So blades are longest at the front (LP) and shortest at the rear (HP)
Mass flow = density × area × axial velocity
long blades at the front
short blades at the back
casing tapers in
air annulus: the space between drum and casing
Front
Rear
Density
1 ×
≈ 12 ×
Annulus area
1 ×
≈ 1/12
Axial velocity
V
V (same)
Convergence by tapering the casing, the drum, or both
If the area stayed the same, the axial velocity would fall
Chapter 16 · Compressors
11
e-books by eatpl.in · © 2026 · Do not copy or distribute
The space between the rotor drum and the compressor casing is called the air annulus. As the air is compressed into a smaller and smaller volume, its density rises. The mass flow is the density times the area times the axial velocity, and the same mass of air passes every row. So if the area stayed the same, the axial velocity would fall as the density rose. To hold the axial velocity constant, the annulus is made to converge from the front to the rear, by tapering the casing, the drum, or both. In this example the density rises about twelve times, so the area at the back is about a twelfth of the area at the front. The visible result is that the blades get shorter from the front to the rear. The longest blades are at the front, in the low pressure compressor, and the shortest are at the rear, in the high pressure compressor.
Part 2 · Axial
Blade angle of attack, and why blades are twisted The blade meets a relative airflow made of two parts: the axial velocity of the air, and the rotational speed of the blade Blades are set at the best angle of attack for the design rpm Blade speed grows with radius, so the relative flow angle changes from root to tip Blades are twisted root to tip to keep the correct angle of attack along the span
The angle of attack comes from two velocities
axial velocity
blade speed
(reversed)
relative
airflow
blade set at about 5°
to the relative airflow
(dashed: chord line)
Why the blade is twisted (8,000 rpm, 150 m/s axial)
root
r = 0.25 m
blade speed 209 m/s
flow at 54° to axis
tip
r = 0.50 m
blade speed 419 m/s
flow at 70° to axis
Tip moves faster, so the blade angle changes root to tip
Chapter 16 · Compressors
12
e-books by eatpl.in · © 2026 · Do not copy or distribute
A compressor blade is a small wing, and like any wing it has an angle of attack. That angle depends on the relative airflow, which is made up of two velocities: the axial velocity of the air coming along the engine, and the rotational speed of the blade moving across it. Put them together in a velocity triangle and you get the direction from which the air meets the blade. The designer sets the blade at a small angle of attack to that relative airflow, at the design rpm. Now look at the numbers below. At eight thousand rpm the root, at a quarter of a metre radius, moves at about two hundred and ten metres per second, but the tip, at half a metre, moves at about four hundred and twenty. So the relative flow meets the tip at a much flatter angle, and the blade has to be twisted from root to tip to keep the correct angle of attack all along it.
Part 3 · Stall and surge
Compressor stall: the triangle goes wrong
Design point
axial 150 m/s
blade
speed
α ≈ +5°
smooth flow
Axial velocity too low
axial 90 m/s
blade
speed
α ≈ +16°
angle too high: STALL
Axial velocity too high
axial 220 m/s
blade
speed
α ≈ -5°
angle negative: choking
Stall = an imbalance between axial velocity and blade speed . Too little axial velocity for the rpm: angle of attack too high, the blade stalls.
Chapter 16 · Compressors
13
e-books by eatpl.in · © 2026 · Do not copy or distribute
A compressor stall is simply what happens when the balance between the axial velocity and the blade speed is upset. On the left is the design point, with the blade meeting the air at about five degrees and the flow smooth. In the middle, the blade speed is the same, but the axial velocity has fallen. The relative airflow swings round, the angle of attack grows to around sixteen degrees, and the blade stalls, just like a wing at too high an angle. On the right the axial velocity is too high. The angle of attack becomes negative, the passage chokes, and the air cannot pass smoothly either. The same thing happens if the rpm changes and the axial velocity does not follow it. So whenever you meet a question on the cause of a compressor stall, think of this triangle: anything that disturbs the relationship between the axial velocity and the rotational speed.
Part 3 · Stall and surge
What upsets the balance Any of these can start a stall, at one blade at first:
Cause What changes Effect on the blade Throttle opened too fast (too much fuel) combustion back pressure rises axial velocity falls, α too high Throttle closed too fast (too weak a mixture) back pressure collapses axial velocity rises, α negative Running above or below design rpm blade speed changes α no longer matched Crosswind, manoeuvres, high AoA, turbulence, intake ice disturbed intake airflow axial velocity falls locally Dirty or damaged compressor less compression axial velocity falls Dirty or damaged turbine less power to the compressor axial velocity falls
Chapter 16 · Compressors
14
e-books by eatpl.in · © 2026 · Do not copy or distribute
Here are the usual causes of a compressor stall, each traced back to the velocity triangle. Opening the throttle too quickly sends a surge of fuel into the combustion chambers; the back pressure rises, the axial velocity falls, and the angle of attack climbs too high. Closing the throttle too abruptly does the reverse: the mixture becomes very weak, the back pressure collapses, the axial velocity rises and the angle can go negative. Running the engine above or below its design rpm changes the blade speed. Anything that disturbs the air arriving at the intake, such as a strong crosswind, violent manoeuvres, a high angle of attack, turbulence or ice on the intake, reduces the axial velocity, at least locally. And a dirty or damaged compressor, or a damaged turbine that cannot drive the compressor properly, lowers the compression and with it the axial velocity. In the exam, if a list contains all of these, the answer is all of them.
Part 3 · Stall and surge
From stall to surge Stall is progressive : one blade, then a stage, then several stages; a partial breakdown of flow Stall signs: vibration rises and EGT rises (less air to cool the gas) Surge : a complete breakdown of airflow, all stages stalled; flow may reverse out of the intakeSurge signs: loud bang , large EGT rise, loss of thrust, possible yaw Action: close the throttle slowly
Stall grows; surge is the end of the road
One blade stalls local breakdown of flow
One stage stalls partial breakdown: STALL
Several stages stall vibration ↑ EGT ↑
All stages stalled: SURGE flow breaks down, may reverse: bang, EGT ↑↑, yaw
Close the throttle SLOWLY
Most surges come from fuel system faults or mishandling
Severe surge can bend rotor blades into the stators
Chapter 16 · Compressors
15
e-books by eatpl.in · © 2026 · Do not copy or distribute
A compressor stall is progressive. In theory it could start at a single blade, spread to the whole of one stage and then to several stages, and if nothing is done it affects the whole engine. While it is a partial breakdown of the airflow we call it a stall. The signs are a rise in vibration and a rise in the exhaust gas temperature, because less air is reaching the combustion chambers to cool the hot gas. If it keeps getting worse, all the stages stall and the airflow through the engine breaks down completely. That is a surge. In a severe surge the flow can even reverse, with air blown out of the intake and a loud bang, a big rise in E G T and a loss of thrust that may make the aircraft yaw. Surges are usually caused by fuel system faults or mishandling. The correct action is to close the throttle of the affected engine slowly.
Part 3 · Stall and surge
Off-design running: front and rear Blades are designed for the best angle of attack near maximum rpm At low rpm the pressure ratio falls, so the air at the rear is less dense and takes up more volume The rear stages choke , holding back the flow; the front axial velocity falls So at low rpm (low pressure ratio) the front stages stall The higher the design pressure ratio, the harder this mismatch is to manage
1
2
3
4
5
6
7
8
9
10
0.5
1
1.5
compressor stage (front → rear)
Axial velocity ÷ design value
choking (α negative)
stall (α too high)
design rpm
low rpm (low pressure ratio)
above design rpm
Low rpm: front stages stall, rear stages choke
Above design: front stages choke, rear stages stall
Chapter 16 · Compressors
16
e-books by eatpl.in · © 2026 · Do not copy or distribute
A compressor is designed to give its best performance near maximum rpm, yet we must be able to throttle it back. Here is what goes wrong. At low rpm the pressure ratio falls, so the air reaching the rear stages has not been compressed as much as the designer expected. It is less dense, it takes up more volume, and the converging annulus at the back is now too small for it. The rear stages choke: the axial velocity there is too high. That choking holds back the flow through the whole compressor, so the axial velocity in the front stages falls, their angle of attack rises and they stall. That is the blue line. So with a low pressure ratio, the stall is most likely at the front. Above design rpm the pattern reverses, the purple line, with the rear stages close to the stall. The higher the design pressure ratio, the worse this problem becomes.
Part 3 · Stall and surge
The compressor map and the surge margin
40
50
60
70
80
90
100
110
1
5
10
15
20
25
30
35
air mass flow (% of design)
Pressure ratio
60%
70%
80%
90%
100%
105%
SURGE / STALL
surge line
working line
rapid acceleration
surge margin
Grey lines: constant
compressor rpm
Too much fuel too fast
pushes the operating point
up towards the surge line
The fuel control limits the
rate of acceleration to keep
a safe surge margin
Stall and surge come from an imbalance between airflow and pressure ratio . The working line keeps a surge margin ; the fuel control limits acceleration so it is not used up.
Chapter 16 · Compressors
17
e-books by eatpl.in · © 2026 · Do not copy or distribute
Designers summarise all this on a compressor map. Across the bottom is the air mass flow, up the side the pressure ratio, and each grey line is one constant compressor rpm. Joining the top of each rpm line gives the surge line. Above it and to the left, the compressor cannot hold that pressure ratio with so little airflow, and it stalls or surges. The engine normally runs along the green working line, and the gap between the working line and the surge line is the surge margin, a built-in safety allowance. Now look at the orange path. If the pilot slams the throttle open and too much fuel arrives too quickly, the back pressure rises before the compressor can speed up, and the operating point swings up towards the surge line. That is why the fuel system has an acceleration control, metering fuel during rapid acceleration to keep a safe margin and prevent a stall or surge.
Part 4 · Control
Variable inlet guide vanes Fitted just ahead of the first rotor stage; they pivot on their own axes, automatically At low rpm they are closed , giving the most swirl in the direction of rotation Swirl corrects the relative airflow to a good angle of attack on the first rotor Closing them also reduces the mass flow at low rpm, matching the front to the rear They open progressively as rpm rises: smooth, rapid acceleration
Low rpm: VIGVs closed to add swirl
Without swirl
α ≈ 16°
stalled
With VIGV swirl
α ≈ 9°
acceptable
axial 90 m/s
air leaving the VIGVs swirls with the rotor
Swirl in the direction of rotation reduces the
relative speed, so the relative airflow swings back
to a safe angle of attack on the first rotor row.
Closed at low rpm, opened progressively as rpm rises
Chapter 16 · Compressors
18
e-books by eatpl.in · © 2026 · Do not copy or distribute
The first line of defence is the variable inlet guide vane, the V I G V. These are fitted just in front of the first rotor stage, on engines that are prone to stall at low rpm or during acceleration and deceleration. Each vane pivots about its own axis, under automatic control. At low rpm the vanes are closed, angled to give the air the greatest swirl in the direction of rotation. Look at the triangles. Without swirl, the slow axial flow meets the blade at about sixteen degrees and it stalls. With the swirl, the air is already moving with the blade, so the relative speed is less and the angle of attack comes back to about nine degrees. Closing the vanes also chokes back the mass flow a little at low rpm, which matches the front of the compressor to the rear. As the rpm rises, the vanes open progressively, which allows a smooth and rapid acceleration.
Part 4 · Control
Variable stator vanes After the first rotor the flow can still go wrong further back Some engines make the front stator rows variable as well as the IGVs As rpm falls below design, they are progressively closed This keeps an acceptable angle of attack on the following rotor row Moved together by unison rings and an actuator scheduled by the fuel control
Variable stators: IGVs and front stator rows
actuator
unison ring and levers
IGV
fixed rear stages
high rpm: open
low rpm: closed
Scheduled automatically by the fuel control (rpm, inlet temperature)
Chapter 16 · Compressors
19
e-books by eatpl.in · © 2026 · Do not copy or distribute
Getting the air safely through the first rotor is not the end of the story. Further down the compressor the flow can still go wrong when the engine runs away from its design point. So some engines also make the first few rows of stator vanes variable. Each vane has a small lever on the outside of the casing, and all the levers of one row are joined to a unison ring. An actuator, scheduled by the fuel control according to the rpm and the inlet temperature, turns the rings, so every vane in the row pivots together. As the compressor speed falls below its design value, the stators are progressively closed, turning the air so that it meets the following rotor blades at an acceptable angle of attack. At high rpm they open again, nearly in line with the flow. The rear rows, where the problem is smaller, are usually fixed.
Part 4 · Control
Compressor bleed valves Fitted at the intermediate (mid) stages Opened automatically at low rpm and during acceleration ; closed at high rpm Dumping air overboard raises the flow and axial velocity in the front stages …and reduces the flow through the rear , relieving choking Any bleed costs thrust , raises EGT and SFC ; a valve stuck closed at low rpm risks stall
Bleed valve: low rpm and acceleration
bleed valve (open)
air dumped overboard
front: more flow, higher
axial velocity, less stall
rear: less flow,
choking relieved
Price of bleeding air
Thrust for a given throttle
falls ↓
Exhaust gas temperature
rises ↑
Specific fuel consumption
rises ↑
The same penalties apply to bleed taken for anti-icing or air conditioning
Chapter 16 · Compressors
20
e-books by eatpl.in · © 2026 · Do not copy or distribute
Bleed valves attack the same mismatch in a different way. At low rpm the rear of the compressor is choking, which starves the front stages of airflow. If we fit a bleed valve at an intermediate stage and open it at low rpm, or during acceleration, some of the excess air escapes overboard. More air can now pass through the front stages, so their axial velocity rises and their angle of attack falls back to a safe value, while less air has to pass through the choking rear stages. As the rpm rises the valves close automatically. There is a price. Any air bled from the compressor, whether to prevent a stall or for anti-icing or air conditioning, reduces the mass flow through the engine. For a given throttle position the thrust falls, the exhaust gas temperature rises because there is less cooling air, and the specific fuel consumption rises. And if a bleed valve sticks closed at low rpm, expect a compressor stall.
Part 4 · Control
Two and three spools Splitting the compressor into two or three sections, each with its own turbine and shaft Compressor + shaft + turbine = a spool Speeds rise from LP to HP: LP slowest, HP fastest The LP compressor is driven by the rearmost turbine ; its blades are the largest On throttling back the LP spool slows more , keeping each compressor near its best angle of attack
Three-spool engine: each compressor has its own turbine
burner
fan
IP comp.
HP comp.
HP
turb.
IP
turb.
LP turbine
three concentric shafts: LP innermost
Spool
Driven by
Speed
LP (fan)
rearmost turbine
slowest
IP
middle turbine
faster
HP
first turbine
fastest
Throttle back: the LP spool slows more than the HP,
keeping each compressor nearer its best angle of attack
Chapter 16 · Compressors
21
e-books by eatpl.in · © 2026 · Do not copy or distribute
Early axial engines got higher pressure ratios simply by adding more stages on one shaft. But all those stages had to turn at the same speed, and as the rpm fell the velocity triangles went wrong and stall became an ever present problem. The answer was to split the compressor into two, and later three, sections, each driven by its own turbine through its own shaft. A compressor, its shaft and its turbine together form a spool. Each spool settles at its own speed, and they get faster from front to back: the low pressure spool turns slowest and the high pressure spool fastest. The low pressure compressor, or fan, is driven by the rearmost turbine on the innermost shaft, and its blades are the largest. Watch the scene as the throttle is closed: the spools slow down by different amounts, which keeps each compressor much closer to its ideal angle of attack, and then they accelerate again.
Part 4 · Control
Active clearance control All stalls come back to the angle of attack , set by axial velocity and blade speed Controlling the annulus at the HP end helps control the axial velocity The gap between blade tips and casing changes as parts heat and cool Cooling air through tubes round the casing makes it shrink Smaller tip gaps: less leakage , better efficiency, better stall margin
Active clearance control
Casing hot: gap opens
casing
tip gap
blade
Casing cooled: gap closes
casing
tip gap
blade
cooling air
Large tip gap: air leaks round the tips,
efficiency falls and the stall margin shrinks
Cooling air blown through tubes round the casing
shrinks it, holding the gap small at the HP end
Chapter 16 · Compressors
22
e-books by eatpl.in · © 2026 · Do not copy or distribute
A more recent development is active clearance control. Remember that every stall comes back to the angle of attack, which depends on the axial velocity and the blade speed. If the axial velocity can be controlled over the whole speed range, the chance of stall or surge falls. One way to do this is to control the size of the annulus at the high pressure end, something once thought technically impossible. The trick is temperature. The casing and the blades grow and shrink at different rates as they heat up and cool down, so the gap at the blade tips changes. A big gap lets air leak round the tips, which wastes work and eats into the stall margin. In active clearance control, cooling air is blown through tubes running round the outside of the compressor casing. The casing shrinks, and the gap between the casing and the blade tips is held small.
Part 4 · Construction
Discs, roots and casings Rotor shaft in bearings, coupled to the turbine shaft to allow small misalignments Centrifugal loads: blades are fixed to discs on the shaft Usual fixing: a dovetail root secured by a pin or locking tab Small engines: blades made integral with the disc, a blisk Casing: aluminium alloy at the front, steel in the middle, nickel alloy at the hot HP end
Rotor: blades on discs on a shaft
shaft
disc
dovetail roots
pin or locking tab
blades
or one piece:
a "blisk"
aluminium alloy
steel alloy
nickel alloy
front, cool
rear, up to about 600 °C
Casing materials change as the temperature rises
Chapter 16 · Compressors
23
e-books by eatpl.in · © 2026 · Do not copy or distribute
Now to how a compressor is built. The rotor shaft runs in bearings and is coupled to the turbine shaft in a way that allows for small variations in alignment. The centrifugal loads on the blades are enormous, so each row of blades is fixed to a disc, and the discs are fitted on the shaft. The most common fixing is a dovetail root, slid into a matching slot in the rim of the disc and secured by a pin or a locking tab. In small engines it is hard to find a practical fixing that does not make the disc too heavy, so the blades and the disc can be made in one piece, called a blisk, from blade and disc. The casing changes material along its length as the air gets hotter: aluminium alloy at the cool front, steel alloys in the intermediate stages, and at the high pressure end only nickel based alloys can stand the temperature.
Part 4 · Construction
Rotor blades and stator vanes Rotor blades: aerofoils, usually forged stainless steel ; titanium in cooler LP stages Dovetail roots are loose when stopped, seated by centrifugal force when running; the slight freedom relieves stress Windmilling on the ground they rattle like a bag of nails Long stator vanes are shrouded at their inner ends to minimise vibration Steel and nickel resist impact and erosion; a heavy rub can ignite titanium
Rotor blade
dovetail root
aerofoil,
twisted
loose when stopped: seated
by centrifugal force when
running
Stator vane segment
inner shroud
stops the long vanes vibrating
casing
Titanium: light and strong, but only where it is cool;
a heavy rub can set titanium on fire
Steel and nickel alloys resist impact and erosion
Chapter 16 · Compressors
24
e-books by eatpl.in · © 2026 · Do not copy or distribute
Rotor blades are aerofoils, normally drop forged from stainless steel and machined to close tolerances, and they get smaller from front to rear to suit the converging annulus. In the cooler low pressure stages they may be titanium. A dovetail root does not hold a blade rigidly. The blades are quite loose until centrifugal force seats them firmly when the engine runs, and that slight freedom of movement relieves stress concentrations at the root. If you turn a compressor slowly on the ground, the blades rattle like a bag of nails being shaken. That is normal. Stator vanes are aerofoils too, fixed to the casing or to retaining rings, and the longer ones are shrouded at their inner ends to minimise vibration. Aluminium vanes coped badly with foreign objects, so steel and nickel alloys are used. Titanium appears in the early stages, but a heavy rub can generate enough heat to set titanium alight.
Part 4 · Construction
Fan blades The fan is the LP compressor of a high-bypass engine Light blades limit the out-of-balance forces if a blade fails Early fans: solid titanium with mid-span snubbers to stop flutter Snubbers add weight and spoil the supersonic flow near the tips Now: wide-chord blades, titanium skins on a honeycomb core , no snubber
Early: solid titanium
narrow chord + mid-span snubbers
snubbers
snubbers add weight and
spoil the supersonic tip flow
Now: wide-chord fan
hollow, honeycomb core, no snubber
titanium skins on a honeycomb
core: strong, light, stable
A light blade limits the out-of-balance load if one fails;
carbon fibre was tried, but was too flexible
Chapter 16 · Compressors
25
e-books by eatpl.in · © 2026 · Do not copy or distribute
In a high bypass engine the low pressure compressor is the fan, driven by the low pressure turbine, and its blades are huge. They must be light, because if a blade ever fails, the out of balance forces on the rotor depend on the mass that is lost. The first fan blades were solid titanium, strong and fairly light. Even so, they were narrow, and to prevent aerodynamic instability they needed a support at mid span, the snubber, sometimes two. Snubbers add weight, and they interfere with the supersonic flow over the outer part of the blade. Carbon fibre was tried, but it was too flexible. The real advance was to build the blade from two titanium skins on a honeycomb core. That gives strength with less weight, allowing the wide chord fan blade, which is stable enough on its own, so no snubber is needed.
Part 4 · Keeping it clean
Contamination and compressor washing Dirt on the blades seriously reduces efficiency and performance, and raises the risk of stall.
Clean blades
smooth aerofoils, full efficiency
Contaminated blades
salt and dirt: rough, misshapen aerofoils
Washing
intake
spray nozzle
motoring
or at idle
Compressor deposits Mostly salt and industrial pollution , spoiling the blade aerofoils.
Turbine: sulphidation Sulphur deposits from the fuel attack turbine blades and NGVs, eroding their finish.
Desalination wash Fresh water sprayed in while motoring on the starter or at idle: removes salt.
Recovery and grit Emulsion cleaner for heavier dirt; abrasive grit (walnut shells) at idle, mainly centrifugal engines.
Chapter 16 · Compressors
26
e-books by eatpl.in · © 2026 · Do not copy or distribute
Finally, keeping the compressor clean. Deposits on the compressor blades, mostly salt and pollution from industrial areas, spoil the aerofoil shapes and reduce efficiency, and that can seriously affect the performance of the engine. In the turbine the contamination is different. It is sulphidation, a build-up of sulphur deposits from the burning fuel, which destroys the shape of the turbine blades and nozzle guide vanes and slowly erodes their surface. If the main problem is salt, a timely rinse with fresh water, sprayed into the intake while the engine is motored on the starter or running at idle, will remove it. That is a desalination wash. Heavier dirt needs an emulsion type cleaner applied the same way, a performance recovery wash, which helps the turbine too. A more vigorous treatment, mostly for centrifugal engines, injects abrasive grit such as broken walnut shells at idle, but it burns in the combustor and does little for the turbine.
Summary
Compressors in one page Centrifugal Eye → impeller (V, p, T ↑) → diffuser (V ↓, p, T ↑). 50 : 50 split, about 4 : 1 per stage, two stages max, cascade vanes in the elbows.
Axial Stage = rotor then stator; 1.1 to 1.2 : 1 per stage; p and T rise in both rows, V rises in the rotor only; up to 35 : 1 or more.
Annulus and blades Annulus converges to hold axial velocity; blades shortest at the rear; twisted to keep α along the span.
Stall Axial velocity and blade speed out of balance. Partial breakdown; vibration and EGT rise. Low rpm: front stages stall.
Surge All stages stalled, complete breakdown, possible reversal, bang, EGT rise, yaw. Close the throttle slowly.
Cures VIGVs and VSVs closed at low rpm; bleed valves open at low rpm; multi-spool (LP slowest); clearance control; acceleration control.
Chapter 16 · Compressors
27
e-books by eatpl.in · © 2026 · Do not copy or distribute
Let us pull it together. In a centrifugal compressor the air enters the eye, gains velocity, pressure and temperature in the impeller, and is slowed in the diffuser, where the pressure and temperature go on rising. The rise is shared about equally, a stage gives about four to one, and two stages is the practical limit. In an axial compressor a stage is a rotor followed by a stator, each giving only one point one to one point two to one; pressure and temperature rise in both rows, while velocity rises only in the rotor. The annulus converges to hold the axial velocity constant, so the blades get shorter towards the back, and they are twisted along their length. A stall is an imbalance between axial velocity and blade speed; a surge is the complete breakdown of flow, and the action is to close the throttle slowly. The cures are variable vanes, bleed valves, multiple spools, clearance control and controlled acceleration.