Bjorn’s Corner: Turbofan engine challenges, Part 5

By Bjorn Fehrm

December 02, 2016, ©. Leeham Co: We will now look at the combustor area in our series on modern turbofan engines. There is a lot of activity in this area, as it sets the level of pollution for the air transportation industry for some important combustion products.

We will also finish off the compressor part of our series by looking at the bleeding of cooling air for the engine and for servicing the aircraft with air conditioning and deicing air.


Figure 1. GasTurb principal representation of a three-shaft turbofan like our reference Rolls-Royce Trent XWB. Source: GasTurb.

The amount of air which is tapped from compressor stages for cooling and other purposes can exceed 20% of the core flow (some of the flow paths are shown in Figure 1).  At that level, it has a marked influence on the performance of the engine.

Compressor bleed air

A turbofan engine has the need for compressed air for many purposes:

  • For cooling of engine parts, static or rotating.
  • For clearance control of rotating parts.
  • For functioning as part of a seal. The air enters with over-pressure on one side of the seal to make sure there is a positive flow of air to stop, e.g., combustion gases from getting into areas which contain combustible oil or oil residues.
  • For deicing of the engine’s front parts.
  • For deicing of the aircraft’s wing and empennage leading edges.
  • For functioning as the air source for the aircraft’s air conditioning system.
  • For starting a second engine via its air starter.

The air is always tapped at the earliest possible place in the compression chain which has the appropriate pressure level. The least amount of energy is then spent on the air and it has the lowest possible temperature.

Figure 2 describes tapping of the air from the main compressor of the CFM56-5 engine for the Airbus A320 series (note the four variable stator guide vanes and the knife edge seals at the beginning/end of the compressor drum and at each vane end).


Figure 2. High-pressure compressor of CFM56-5A with tapped air flows. Source: CFM. Click to see better.

The CFM56-5 has up to three active clearance control systems:

  1. For the high compressor rotor (system called RACC).
  2. For the High-Pressure Turbine (HPT)
  3. For the Low-Pressure Turbine (LPT).

All these mix colder air from, e.g., the booster area with hotter air from different parts of the compressor and routes the mix to the rotor drum (see arrows below the compressor disc ends) or casings that should be temperature-manipulated.

Controlling the temperature of the drum or casings controls the mechanical expansion of the metal and achieves the appropriate tip clearances for the blades of the section.

As can be seen in Figure 2, air is tapped from several places on the compressor. This is easy to achieve because each compressor stage can be opened to the outer casing at several places around the compressor.

Figure 3 shows a compressor which has holes leading to casing pockets with outlets A and B at the sixth and ninth stage of the compressor (air is tapped from one or the other dependent on compressor RPM and therefore pressure level).


Figure 3. Compressor cross section (lower part) with stator seal back-flow (red arrow) and A and B taps of compressed air. Source: Rolls-Royce “The Jet Engine” and Leeham Co.

In Figure 2, one can see that air enters the rotor drum by a gap at the first stator vane. This air is then mixed with air from the fifth stage under the control of the RACC valve. Air is tapped in many more places. For the CFM56, air is even tapped at the beginning and end of the combustor, in a cold area pockets, for cooling of the high-pressure turbine stator and rotor. Figure 1 shows some main paths for secondary air in an engine.

Bleeding air lowers the performance of the compressor and hence the engine. When bleed air is tapped for cabin air conditioning and de-icing, it increases the fuel consumption of the engine by up to 5%.

The bleed air tapped from the high-pressure compressor has to high temperature to be routed around the aircraft. It’s therefore pre-cooled in a pre-cooler which uses the fan’s airstream for cooling. Figure 4 shows the bleed air plumbing and pre-cooler for the GEnx-2B for the Boeing 747-8i compared to the same engine (in GEnx-1B variant) when used in the 787 (which does not use bleed air for aircraft systems).


Figure 4. GE’s GEnx-1B for Boeing’s 787 (left) and GEnx-2B 747-8i (right) with bleed air pre-cooler. Source: Boeing.


Modern combustors are made as rings with fuel injectors and combustor cans housing the flames, Figure 4. The air is slowed down to around M0.2 at the entrance to the combustor by the divergent entry nozzle seen in the figure. The low axial speed of air is necessary to guarantee that the flame is stable and that, e.g., a re-light of the engine can be reliably done while in the air.


Figure 4. GE9X combustor with lean burn TAPS pre-swirl fuel injectors. Source: GE Aviation.

The combustor is an area where a lot of research is done. The high compression ratio of modern engines creates a potential high level of NOx emissions. Research programs like the European ACARE aims to define combustors with low levels of NOx, Figure 5.


Figure 5. Graph from a Rolls-Royce presentation on engine research. Source: Rolls-Royce.

The emission of CO2 is 100% coupled to the fuel consumption of the engine. With lower Specific Fuel Consumption (SFC, consumption per unit of thrust), the level of CO2 emissions are reduced accordingly.

This is not the case with NOx. Reduction of NOx levels require the division of the combustion into a rich pilot flame area (for reliable function over the power spectrum) and a lean burn area for low NOx generation.

The different OEMs have different programs to achieve low NOx combustors. Figure 6 is from a presentation by the Rolls-Royce-affiliated Loughborough University in the UK. It shows how combustors develop from classical types (lower left) to new types with clearly divided flame areas of rich and lean combustion.


Figure 5. Research combustors with changed combustion into rich and lean areas. Source: Loughborough University.

Here are the design requirements according to Loughborough for a modern combustor:


Figure 6. Combustor requirements according to Loughborough University. Source: Loughborough University.

The term AFR stands for Air Fuel Ratios. “Good temperature traverse quality” will guarantee that the exit nozzle from the combustor, which acts as the inlet nozzle for the high-pressure turbine, has a gas flow with an even temperature distribution so that the nozzle and turbine blades will not be subject to “hot spots.”

These parts of the engine are living a life on the verge of melting from temperatures as high as 1,700°C. Any “hot spots” on top of the high average temperature would kill the life of the components. More on turbines in the next Corner.

Note also the words “thermo-acoustic properties”. Resonant acoustic buzz is a major problem in combustors as it means high frequency variable loads on the structure in the combustor. This leads to fatigue cracks and the breakdown of e.g. the combustor’s can insert.


13 Comments on “Bjorn’s Corner: Turbofan engine challenges, Part 5

  1. Interesting as ever. Regarding this point.
    Bleeding air lowers the performance of the compressor and hence the engine. When bleed air is tapped for cabin air conditioning and de-icing, it increases the fuel consumption of the engine by up to 5%.
    If bleed air is not used then electrical power will need as a replacement. This will require larger generators which will take more shaft power to deliver the increased electrical power – so the overall change in efficiency of the engine as an overall system is nowhere near 5%.
    These are as ever with aeronautical issues a matter of design trade offs.

    • Hi Tim,

      the 5% was not mentioned as a difference to a more electrical alternative. When SFC (or really TSFC) is given for airliner engines it’s normally without power (i.e. eletrical generators and hydralic pumps) and bleed off-take for aircraft services. It’s done this way to not have the implementation affecting the SFC. But the real installed SFC is degraded by nacelle losses (the values are test stand values) and the different power/bleed off-takes. And the effect on the actual SFC in the aircraft during flight is large, close to 10% with bleed and power off-takes making up around half of that.

      The debate on bleed (i.e. A350) or more electrical implementation (i.e. 787) is the one about which conversion chain is the more efficient; engine compressor + pre-cooler + air condition unit vs. electrical alternators + electrical air condition compressors + air condition unit. The latter is slightly more efficient but more complicated to implement because it’s new and the high electrical power requirements forces liquid cooled power conversion electronics. Once the power converters get more efficient and can be air cooled, we will see more aircraft choosing no aircraft bleed air.

      • Bjorn:

        Thank you, that is as clear and concise an explain of the difference I have see.

        the one area I think that eclectic has the advantage is in routing flexibility.

        At work, Ducts take precedence as they are large and not easily changed for routes (and lose a lot of efficiency if they are)

        theoretically it would seem that you free up a lot of structural design trade offs consideration if you don’t have to ensure the opening for the tubes.

        Heat is heat and its easier to get rid of out on the wing I gather?

        Can you design a better wing though if you don’t have to make room for heat exchangers and tubes runs or is that a non issue in the overall?

        • I know the potential leak of hot air (air temperature after the pre-cooler is around 200°C) is a problem with bleed systems. Aluminium or CRFP does not like 200°C so you need leak detectors along the routing. The tubes will take place but if you go for a more electrical system you have very high Voltage and current routing to do for e.g. the de-icing mats. And you need to route high power lines from the APU to the engines for engine starting instead of the bleed air tube. You also need to house the power conversion electronics, the 787 has an extra equipment bay in the belly for this reason.

          So yes, not having to route the air tubes is an advantage (both for leak and place reasons) but your electrical routing problems gets augmented instead. The trend is clear, we will go more electrical, it has advantages. But there is so much that has to follow in that change. One example; all ground power units in all the world’s airports are laid out to supply air for an engine start + normal level electrical power. With the 787 you need no air for starting but much more electrical power, so you need three units for an engine start (if you can’t use the aircraft’s APU for some reason). The change will take time.

          • Bjorn: Thank you

            We supply tow GPUs for our 777s, though not due to power needs (not on the freighters)

            Its a split bus on 777 and Boeing won’t allow a Y splitter even though the power load is well below a 90KVA unit .

            Pretty funny as internally our Electronic GPUs are nothing more than a contactor splitter (both contactors hooked to the AC supply source)

            It is an issue when one goes down and thy have supply two engine driven units.

            I have 600 KW generators I take care of.

            I think, sheese, 787 has the equivalent of two of those 900 hp beats (1.25 MW if I remember right) split between the engines and the APU.

            I can power our whole facility with 1.2 MW, lights, computes, conveyors, fans, AC, plug in unit and those GPUs.

  2. Bjorn,

    I am not 100% sure that the “bleedless” approach of the GEnx-1B is going to prevail. No other engine maker (RR, P&W, Russians) and no other airframer (Airbus, Embraer, Bombardier) has embraced that approach. Boeing also did not choose it for its 747-8 and 777X (perhaps since these are existing designs and would require unacceptable structural modifications?). You do lose efficiency since you have to first convert shaft power to electricity, then transmit it and then use it in electrical heaters. The overall efficiency advantage may not be that compelling, at least according to some statements from RR. Their new designs such as the Ultrafan are not “bleedless.” You still need to bleed air for turbine cooling etc. so that “bleedless” is sort of misleading terminology. Correct me if I am wrong.
    Also there is a small typo: knife egg should be knife edge.

    • The Trent 1000 is also bleedless.

      The requirement for it is at an aircraft and engine overall system level. A bleedless engine cannot connect to an aircraft that is not designed for it and vice versa.

      My guess would be the 777X requires too much change for a bleedless engine to be incorporated.

      The more interesting point would be why the 787 has it and the A350 doesn’t…

      • The A350 was in a position to inherit a lot of modern systems from the A380, among these an electronically controlled bleed systems. Airbus could focus getting the aircraft right while the systems and avionics/FBW was to a large degree taken over from A380 with updates. Boeing had their last design, the 777 which had a larger gap to the 787, it inherited the systems concept from the Sonic Cruiser which was a more electrical approach. Hence both the GEnx-1B and Trent 1000 have no supply of bleed to the aircraft via a pre-cooler.

      • My bad! Forgot the Trent 1000. That once again begs the question: Why did RR decide not to go “bleedless” in their future engines to appear decade(s) from now? What do they know that I don’t?

        • RR has not declared whether the Advance/Ultrafan has customer (i.e. aircraft services) bleed or not. They don’t have to. Customer bleed is not a major part of total bleed in the engine. The conversion from Trent 1000 (no customer bleed but two 250 kVA alternators instead) to Trent 7000 (customer bleed for the A330 and one 115kVA constant drive generator) was no big effort for RR ( I dicussed it with them). It required adding taps and remapping the FADEC. If you take the power as bleed air or mechanical power is not that big a change for the engine. Nacelle inlet de-icing and the engines own multiple bleed flows are larger than the customer bleed flows. It has large consequences for the aircraft but not for the engine.

          • Bjorn ; Thank you, makes sense and per some of what I felt about Boeing, they really jumped into the deep end of the pool. To me its amazing the more electric systems did not bit them far worse than they did.

            I would like to add, Building control system (dampers for outside air, valves for heat, dampers for room control boxes called VAV) used to use Air (pneumatics aka bleed air) as a power source.

            They most efficient system was when we did a hybrid where we used electronic controllers but converted the electrical damper signal via a transducer into a Pneumatic action.

            Lovely, simple, easy to trouble shoot.

            Only worked if you had an older building.

            Now they run power to all of the equipment and do it with electro actuators (impossible to trouble shoot, intermittent a pain, very expensive)

            I have one room where they put 22 control transformers on a wall to power all the room control boxes in that part of the project.

            Air tubes would be simpler!

            Downside is you have to have a good air plant.

            Not a problem as Quincy makes a very solid reliable one and with 2 compressors never a problem.

      • Orthogonality ( 3 independent power systems : 2H 2E _and_ 2 P(neumatics).)

        Current power conversion system efficiency and reliability limits. ( connected high losses induce heat cycling which increases fatique and thus reduces service life and general reliability.)

        IMU the 787 application was one maybe two generations too early. Earth bound power conversion show fast growth and makes big steps from a wide innovations base.
        ( look at early digital processors in Airbus FBW: they were taken from well established and optimized lines. nothing newfangled from the electronics industry.)

  3. Oddly the electronics seems to be working very well.

    I too thought it was a Bridge To Far (ref WWII, operation Market Garden)

    I know they have had to revamp some of the end units but the mains system seems to work well.

Leave a Reply

Your email address will not be published. Required fields are marked *