A Quick Explanation of Jet Engines

[Collimatrix]

I've had an idea for a post like this rattling in my head for a while.  I thought we should have some sort of resource and discussion on jet engines, how they work, and what the various figures of merit are.

 

For simplicity, I'll talk about thrust turbines in aircraft, although there are a whole host of important gas turbine applications, like the power gas turbines used in helicopters, tanks and ships, to say nothing of steam turbines.

 

 

This is a representative schematic of a turbojet.  For some reason, all schematic diagrams of turbojet engines show axial compressors, even though centrifugal compressors are still common.

 

Air flows in through the inlet. Rotating compressor blades, attached to a central shaft or spool and stationary stator blades compress air.  This compressed air is fed to the combustors, where fuel is burned to heat the air (in principle heat energy could be added by means other than combustion, as in a nuclear gas turbine).  This hot air is then fed to the turbine section where it spins turbine blades, which are mounted to the same shaft as the compressor blades.  Some of the energy from the hot gas is used to turn the turbines, but the rest rushes out the back where a nozzle extracts maximum thrust from it.

 

For best performance, the pressure ratio, that is, the ratio of the pressure of air in the final stage of the compressor to free atmospheric air, should be as high as possible.  Pressure ratio is related both to specific power (how much thrust you can get per kilogram of engine) and efficiency (how much thrust you can get per kilogram of fuel):

 

 

This chart shows the idealized relationship between efficiency (Y axis) and pressure ratio (X axis).  Higher is better, but after a certain point there are somewhat diminishing returns.

 

Pressure ratio has risen dramatically since the development of the first gas turbines.  The Jumo 004 in the ME-262 had a pressure ratio of 4:1 or so, while the Trent 900 in the Airbus A380 manages nearly 40:1!  Civil turbine engines can support somewhat higher compression ratios, since airliners cruise for long periods of time at a given altitude and speed.  Military aircraft have fighter jocks shoving the throttles this way and that and operate over a wide range of altitudes and speeds, so there needs to be some more wiggle room.  Most contemporary fighter aircraft engines have pressure ratios between 20:1 and 30:1.

 

Additionally, contrary to what the diagram above shows, in most thrust jet engines, not all of the air from the compressor goes through the core:

 

 

A portion of the air bypasses the core, and simply joins the air that went through the core in the nozzle.  The bypass ratio is the ratio (in mass per second) of air that goes around the core relative to the air going through the core.  An engine with any amount of bypass air is called a turbofan, while one without is a turbojet.

 

This bypass air serves two purposes.  The first is that because it did not go through the core and past the combustor, it is relatively cool and therefore helps to cool the engine nozzle.  The variable geometry nozzles (overlapping metal flower looking thingies) used on modern fighter aircraft require some amount of bypass air to maintain structural integrity:

 

 

The second purpose of bypass air is to generate thrust.  At low altitudes and airspeeds, bypass air is a more efficient way to generate thrust than core air.

 

Airliner engines were initially turbojets, but these quickly sprouted bypass ducts.  Airliner engines currently have much higher bypass ratios than fighter aircraft engines, as again, fighter aircraft are expected to operate at least occasionally at speeds and altitudes where high bypass is not efficient.  Additionally, higher bypass ratios make engines wider, and therefore draggier, as well as less responsive to throttle input.  The Trent 900 has a bypass ratio of 8.5:1, while the RB199 has the highest bypass ratio of any fighter engine at 1.1:1.  The F119 in the F-22 raptor has an unusually low bypass ratio of .2:1, making it practically a pure turbojet with just enough bypass air for nozzle cooling only.  It is also possible to make variants of engines that incorporate existing core designs, but have a different bypass ratios.  The F135 in the JSF, for example, has essentially the same core as the F119, but has a bypass ratio of .55:1.  Because the JSF does not supercruise, it is worth trading off some performance at supersonic speeds in favor of efficiency in subsonic.

 

The turbine section of the engine is the most demanding in terms of materials science.  The higher the temperature at which the turbine operates, the more efficient and power-dense the engine can be.  The demands of higher and higher performance have long since pushed temperatures past the operating range of steels, so now turbine blades in top of the line engines are made of magical nickel alloys with magical microstructure, cooling air actively circulated inside of them, and magical ceramic coatings:

 

 

The expertise to make high performance gas turbine blades is kept secret and restricted to a handful of companies in the world's richest nations.

 

Finally, the hot air from the turbine and the bypass air (if any) are mixed and expelled out of a nozzle.  Because they operate within a narrow band of altitudes and speeds, simple, fixed-geometry nozzles are acceptable for airliners.  Combat aircraft typically need complex, actively cooled variable-geometry nozzles.  The most recent development in this area has been the so-called 2D, or rectangular nozzle, as seen in the F119 engine:

 

 

These nozzles are heavier, and sacrifice some efficiency (not that it really matters; the F119 is a mighty beast with thrust to spare) in exchange for lower radar and infra-red cross section.  Additionally, the F119 can point the thrust up or down by twenty degrees, which enhances the agility of the F-22, especially at high altitudes.

 

Most combat aircraft engines feature afterburners.  An afterburner dumps additional fuel into the hot air aft of the turbine stage.  Afterburners provide a dramatic increase in thrust, usually doubling it, but have a disproportionate increase in fuel consumption, usually increasing it by a factor of seven or so.

[Sturgeon]

Question: How are bypass ratios measured?

[Collimatrix]

Empirically measured?  I don't know.  Magical instrumentation, presumably.

 

The numbers themselves refer to mass flow, so the ratio of kilograms of air per second through the bypass divided by the same through the core.

[Xlucine]

The microstructure isn't that complex - creep is the diffusion of atoms which moves grain boundaries, letting the material run away from the load, so designers avoid grain boundaries perpendicular to the applied stress. They used to do this with columnar grains by setting up a temperature gradient along the blade (so all the grains run top-to-bottom), now they do something cool to make the structure a single grain.

 

 

Empirically measured?  I don't know.  Magical instrumentation, presumably.

 

The numbers themselves refer to mass flow, so the ratio of kilograms of air per second through the bypass divided by the same through the core.

 

Maybe a hot wire sensor?

[Toxn]

How much can you lower the materials requirements of the turbine blades by including bypass air in the engine core?

 

I'm thinking of a setup where your combustion chamber consists of a cavity with a number of burner cans suspended inside. Part of the airflow from the compressor then bypasses the cans and cools the turbine blades. This might also be useful as a means to simplify the engine by removing the bypass fan entirely.

[Xlucine]

So some of the air from the compressor goes around the burner cans, then rejoins the exhaust from the cans to meet the turbine, lowering the average temp of the air passing through the turbine? Eeh, you'd get the same effect with less fuel going into the cans and it's bad for brayton efficiency. Burner cans don't have to be stoichiometric, they're pretty far on the lean side

[Toxn]

Good to know.

[Collimatrix]

How much can you lower the materials requirements of the turbine blades by including bypass air in the engine core?

 

I'm thinking of a setup where your combustion chamber consists of a cavity with a number of burner cans suspended inside. Part of the airflow from the compressor then bypasses the cans and cools the turbine blades. This might also be useful as a means to simplify the engine by removing the bypass fan entirely.

 

 

So some of the air from the compressor goes around the burner cans, then rejoins the exhaust from the cans to meet the turbine, lowering the average temp of the air passing through the turbine? Eeh, you'd get the same effect with less fuel going into the cans and it's bad for brayton efficiency. Burner cans don't have to be stoichiometric, they're pretty far on the lean side

 

Toxn's idea wouldn't have exactly the same effect as just burning less fuel.  It would cause a hit to efficiency, but you would at least preserve some of the power density.

 

Actually, gas turbines sort of do what you suggest, just a little differently and without the use of bypass air.  There is a gap between the burners and the turbines, which gives the heated gas some space to cool down.  Since this is adiabatic cooling (or damn close to it), very little energy is lost.  This does cause a bit of a hit to efficiency, but it's necessary to keep the blades from melting.

 

 

The other way that the blades are cooled is that they've got these little networks of holes in them that air (I think it's compressor bleed air, which is essentially the same thing as bypass air) that cool air is actively pumped through.  Given the holes in the surface, I believe that this cooling air joins the hot air going through the turbine.  However, the volume of air is relatively small, so the hit to efficiency is negligible (A GE-90 in a new 747 has a mass flow rate of 1350 kg/sec of air through the bypass and core at cruise setting.  I have difficulty imagining 1.35 metric tons of air.).  Essentially, it's easier to blow air to blow through the turbine blades to cool them off than it is to cool off the gas that's making them hot in the first place.

 

The microstructure isn't that complex - creep is the diffusion of atoms which moves grain boundaries, letting the material run away from the load, so designers avoid grain boundaries perpendicular to the applied stress. They used to do this with columnar grains by setting up a temperature gradient along the blade (so all the grains run top-to-bottom), now they do something cool to make the structure a single grain.

 

 

 

 

The structure is simple, yes, but the tricks you need to use to get metal to form like that are not.

[Toxn]

Yeah, single crystal grain structure is black magic. From what I remember, it involves really fine control of alloy composition and temperature during (what would normally be) the tempering/annealing steps.

 

edit: nope, it's actually really fine control of alloy composition and temperature during the cooling step of a casting process.

[Sturgeon]

That's fascinating, but all I could think of while reading about single-crystal grain structure is how many people would just equivocate details like that and try to come up with alternate history scenarios where Germany has F-135s by 1944, or something.

Which leads me into a rant about the historical explanation for technologies emerging at different times of "no one had thought of it previously"... But I'll say that for another time.

[Collimatrix]

 

Next-generation fighter aircraft engines are likely to feature variable-cycle bypass ratios.

 

A thrust turbine works by accelerating the air that enters it from in front of the aircraft.  A given amount of thrust can be produced either by speeding up a large mass of air a little, or by speeding up a small mass of air very quickly.

 

For a given amount of thrust, it takes more energy to accelerate a small amount of air slowly than to accelerate a small amount of air quickly.  However, as the aircraft goes faster and faster, engines that accelerate a large amount of air slowly produce less and less net acceleration of air relative to the aircraft's velocity, and thrust begins to drop off.

 

High bypass turbofans accelerate a larger mass of air slowly, while turbojets accelerate a smaller mass more quickly.

 

In a variable bypass engine, magic is used to change the amount of air bypassing the core.  This allows the engine to have, within a range, the appropriate bypass ratio for the aircraft's speed.  This should allow the aircraft to have some of the best of both worlds (and some of the worst of both too), and improve efficiency over a wide range of airspeeds.

 

The YF120 engine that was trialed in the YF-22 and YF-23 featured this technology, and the definitive engine for the second-tranche PAK-FA is supposed to as well.  GE is working on the ADVENT seen above, and it is likely that other variable bypass ratio engines will emerge in the future.

[Collimatrix]

 

Thrust figures for jet engines are quoted from ground-based test stands.  The amount of thrust they actually produce when they're living in airplanes is... more complicated.

 

 

This isn't an actual installed thrust chart; I couldn't find one of those.  This is from a flight simulator.  The shape of the curve is probably reasonably close to reality, although the exact numbers perhaps are not.

 

The exact amount of thrust a jet engine can produce is a function of altitude, airspeed, and intake design.

 

At higher altitudes the engine produces less thrust, being air-breathing.  However, there are temperature and drag considerations, so optimal performance may only be possible at substantially higher than sea level.

 

The faster and faster the aircraft moves, the more air is being crammed into the intakes.  This creates drag, but it also increases engine performance.  A well-designed intake will minimize the drag effects and maximize the improvement in performance.

 

At supersonic speeds the intake will generate shock waves.  The shock waves slow down the air entering the engine (which is necessary to avoid breaking the compressor), which also increases the pressure.  Complex, multi-shock and variable geometry intake designs are required to work well in this regime.

[Belesarius]

Article on next gen engines for the US:

 

http://www.defensenews.com/story/defense/air-space/2015/10/10/pratt-ge-battle-future-military-engines/73201826/

[Collimatrix]

Article on next gen engines for the US:

 

http://www.defensenews.com/story/defense/air-space/2015/10/10/pratt-ge-battle-future-military-engines/73201826/

 

Thanks for posting that.

 

The writer is a little confused on terminology and technology, sadly.  A "spool" is a rotating, concentric shaft, not a stream of air.

 

And it's about goddamn time someone started applying all the fancy new technology to turboshafts.

[LoooSeR]

What is this?

 

 

 

[Priory_of_Sion]

Looks like 2D thrust vectoring flaps for an F-15. I don't know why they are in the middle of the nozzles, maybe for simplicity? 

[Collimatrix]

It does look like the F-15S/MTD, which was a proposal to put 2D thrust vectoring nozzles on the F-15, but I don't think that's what it is.  As Priory points out, the nozzle design is different.  Actually, it's so different I'm not sure how it works.  In a typical 2D nozzle you need two paddles, because the angle between them controls the expansion ratio of the nozzle, and you need to be able to adjust that.  It might be a square ejector nozzle.

One thing I notice is that in the topmost picture, the F-15 is shown without cropped wingtips, in the first and second to last picture the horizontal stabilizers are shown without the dogteeth, and in the last picture the horizontal stabilizers are not on booms.

 

Those were all things that were modified or added to the design of the F-15 fairly late in the design.  So I don't think this is an idea for a modified F-15, my guess is that it's from during the F-15's development when they were still playing around with different configurations.  I guess they looked at 2D nozzles at that time.

[LoooSeR]

The first NK-32 engines for Tu-160 bomber will be build in 2016

 

BMPD blog qoute Russian Deputy Defense Minister Yuri Borisov

 

 

   Deputy Minister of Defense of Russia Yury Borisov, during his visit to PJSC "Kuznetsov", said "today in Samara in addition to the state defense order for the repair of engines for strategic aircraft NK-32, NK-25 and NK-12 are preparing for manufacturing of the NK-32 with improved performance. Already this year, the Defense Ministry is to get five of the initial batch of engines in accordance with the signed contract. "  

 

 

   In turn, the plant throughout the year will rebuild test stands for production of aircraft engines NK-32, noted the press service of the company. Until the end of 2016 will be put into operation new galvanic production corresponding to the world's technical and environmental standards, as well as modern standards of safety. The total project cost is estimated at 1.773 billion rubles. In addition, in February this year, completed the reconstruction of the test stand №9, construction and installation work is scheduled to end in april, on the test stand №1. 

[Belesarius]

So those Tu-160 rumors that I posted a while back weren't totally wishful thinking. Interesting.

[Collimatrix]

So those Tu-160 rumors that I posted a while back weren't totally wishful thinking. Interesting.

 

It opens the door to it, but this is more a response to the aging stockpile.  Russian engines typically don't match American ones for flight hours before overhaul, and I don't think the NK-321 is an exception.  Strategic Aviation was running out of fresh engine parts to keep the TU-160 fleet flying.  All the fighting in Syria just accelerated the problem.

 

Even worse is the AN-124, because the engines that uses were made in Ukraine.  I know the Russian aviation industry is working on a replacement, but I'm not sure exactly what that is.

 

NK-321 is, if wikipedia is to be believed, the only other three spool purpose designed combat aircraft engine besides the RB-99.

[Belesarius]

It opens the door to it, but this is more a response to the aging stockpile.  Russian engines typically don't match American ones for flight hours before overhaul, and I don't think the NK-321 is an exception.  Strategic Aviation was running out of fresh engine parts to keep the TU-160 fleet flying.  All the fighting in Syria just accelerated the problem.

 

Even worse is the AN-124, because the engines that uses were made in Ukraine.  I know the Russian aviation industry is working on a replacement, but I'm not sure exactly what that is.

 

NK-321 is, if wikipedia is to be believed, the only other three spool purpose designed combat aircraft engine besides the RB-99.

Yeah, but when we were discussing them building more, IIRC, you said that the biggest obstacle was the # of engines they had for them.  If they are building the engines then building more, if they still have the tooling, isn't as difficult.

 

Totally agree with you that the flight hours from the Syrian fighting has accelerated the timetable for needing more of the engines tho.

[Collimatrix]

Good point.  I guess we'll see if they scrape up the money for them.

 

I wonder if B-1 or B-2 production could be restarted.  I'm guessing not without enormous expense.

[Belesarius]

Well, unlike the Russians who like to keep tooling for everything, the US tends to get rid of tooling as soon as production stops.

[Domus Acipenseris]

https://www.youtube.com/watch?v=zCAR4wPrwrg

 

Video showing Safran engine production in the 1980's.