Bjorn’s Corner: Turbofan engine challenges, Part 3

By Bjorn Fehrm

By Bjorn Fehrm

November 12, 2016, ©. Leeham Co: In our trip through a modern turbofan airliner engine and its technologies, we looked last week at the engine intake and the fan. We now continue with the compressor parts.

As compressors and turbines use the same principles (but in opposing ways), we will look at these principles this week and how their roles in the engine create their special characteristics.


Figure 1. Stylistic cross section of a three-shaft turbofan with section numbers. Source: GasTurb.

As before, to make things concrete, we use a GasTurb simulation of a Rolls-Royce Trent XWB 84k engine to look at practical data when needed. As before, I have no specific knowledge about the engine and will not use any data outside what is public information.

The GasTurb cross section of a three-shaft turbofan is shown in Figure 1. We will examine the sections between station numbers (22) and (3) and (4) and (5) in the general discussion of compressors and turbines. We will then look at some data for common compressors.

Compressor and turbines

Compressor and turbines use the characteristics of a gas that we showed in Figure 2 last week. I repeat it here with an adaptation for this Corner, the beginning of the combustor is changed for a compressor section in the right-hand piece. The compressor rotor consumes axle hp to increase the speed of the air for the following stator vanes to decrease the speed and gain pressure (and temperature).


Figure 2. Use of divergent and convergent ducts to control air speed and pressure. Source: Leeham Co modified RR Jet engine graph.

Most speed increase and decrease is in the tangential direction. The axial velocity over a combined rotor and stator pair is roughly constant, Figure 3.


Figure 3. Section through a compressor with pressure and velocity graphs. Source: Rolls-Royce: The Jet Engine.

The compressor has the more challenging task. The shaft hp is used to sweep the incoming air with a rotor blade which acts like an airplane wing and turns the air in the direction of the rotation. The airspeed and angle to the blade is the vector sum of the incoming air speed and the rotor blade’s local radial speed.

Like for an airplane wing without slats, the angle of attack can only be around 10° max before leading edge separation occurs. Air passing over the blade is going against an increasing pressure gradient, i.e., with a typical pressure gain of 1.3 times over a stage, the pressure in the stator behind the rotor is 30% higher than before the rotor/stator pair. The flow against an increasing pressure makes the trailing edge experience flow separation if the speed of the air is not kept up. Flow separation hurts the diffusion and finally makes the blade stall over its complete surface.

The fact that the blade is turning towards the incoming air from the stator makes the angle of attack increase when the airflow reduces with reduced rotor RPM. With a limited angle of attack capability of the blade, special means are required to let the rotor reduce speed from the design RPM.

These conditions make the compressor fickle. When things don’t fit, the compressor will lose function and can stall violently. The high pressure combustor gas can then go out the front of the engine, a so called “Burp” or surge stall. Alternatively, the compressor can stall for a longer time with little surge action, a so-called rotating stall. Then one needs to shut down the engine to recover normal function.

We will cover what is done to handle the fickle compressor to make it usable over the wide range of conditions which are necessary for an engine in the next Corner. Today we will explain why the turbine, which is using the same principle, is such a robust companion (aerodynamically at least).

The turbine

The turbine is using high energy combustion gas to drive its rotor, Figure 2. Here the gas is going from a higher pressure to lower pressure after the stator/rotor pair. The incoming gas to the turbine blades is turned by the stator to tangential direction and the speed is increased through convergence. The gas is then pushing the blades away from the stream, which is lowering the angle of attack at lower rotor RPMs.

The gas therefore finds it easy to curve around the blade (also at lower RPM) and is meeting a lower pressure on the trailing edge of the blade than at the leading edge. The blade boundary layer is sucked off the blade and the gas speed is kept up. The actual curving is what creates shaft hp (momentum change). A total direction change (deflection) of 90° or more can be achieved with turbine blades. In a compressor rotor the direction change is less than half of that. At the end of the stage the direction of the flow is back to essentially axial, to be ready to enter the next stage.

A turbine stage (stator/rotor) is consequently robust and can easily achieve shaft hp from a 4:1 pressure loss with good conversion efficiency. It can also give a higher shaft hp gain (by virtue of a higher pressure loss) with less efficiency or lower gain with less pressure loss. Whatever you want, it doesn’t complain, just works more or less efficiently. This is why a 20 to one, 10-stage compressor can be driven by a two-stage turbine.

Typical performance values for compressors

Historically, the stage gain of compressors were around 1.2 to 1.25 times per stator/rotor combination. A typical 10 stage compressor delivered a total gain at max RPM at Top of Climb (ToC) of 8-10:1 pressure ratio. Modern compressors have gains of 1.3 to 1.35 times per stage. The new record is a high pressure compressor with a gain of 27 over 11 stages (GE9X). This means a stage gain of 1.35 per stage.

Our Trent XWB has a similar stage gain in its high pressure compressor. The combined stage gain over the Intermediate Compressor and High Compressor is around 40 at ToC, which gives an average stage gain of 1.30 with eight plus six stages in total.

Equally important to the per stage gain is that this is achieved with a high conversion efficiency, shaft hp to pressure gain. The conversion efficiency for compressors used to be between 80-90%. New designs achieve over 90% at their optimum design point (which is designed to be cruise). Values of 92%-93% have been publicized in specialist papers.

Turbines conversion efficiency is typically the same or slightly higher than for compressors, it depends on how many shaft hp they produce (with the corresponding pressure drop). The more you ask them to do, the less efficient they get, more on that later in the series.

Pressure and temperature

Pressure and temperature change go together. An increase in pressure is accompanied by an increase in temperature and vice versa. For air the relationship is: temperature change = pressure change^0.29 (the exponent varies a bit for different pressures and temperatures but is around 0.29). For a combustion gas the exponent is lower, around 0.23. This means that one can measure the one when one want to find out the change of the other, e.g. temp before and after a stage when one want to understand the pressure change.

This also means that the pressure gain in the compressor is accompanied by a temperature increase. At the end of the compressor this forces the material choice to expensive nickel alloys. The temperature drop over the turbine stages allows later stages to be un-cooled.

Next Corner

In the next Corner, we will dive deeper into what is required to have compressors that can work over a wide range of engine RPM’s, air densities and the resulting throughput of air. We will also look at how compressors are constructed and why radial compressor are used for smaller engines.

6 Comments on “Bjorn’s Corner: Turbofan engine challenges, Part 3

  1. looking forward to hearing about why you would choose a radial compressor over an axial. I had always assumed it was a technological generation and cost trade, but there must be more to that as some very new engine designs are still radial compressor based.

    • Sir Stanley Hooker said that their (RR) axial Avon took as many years to reach success as the centrifugal Nene had taken months.

    • The choice between radial and axial compressors depends on the operating point, in particular the non-dimensional flow (mdot*sqrt(T0)/p0, non-dimensional rotational speed (N/sqrt(T0)) and the pressure ratio (p01,p02).

      Here, mdot = mass flow, T0 = total temp, p0 = total pressure and N = shaft speed.

      Note that temperature and pressure need to be total quantities and not static (google for difference, a little too much sidetrack for a post here, wikipedia has excellent article on subject).

      Again, tthe choice between the two types (and any intermediate geometries like a diagonal design, which while rare exist) depend on what pressure ratio is needed at what non-dim flow and non-dim shaft speed. There will be typical curves for each type for the designer to consult, but usually the choice is quite obvious: a radial design cannot handle as large mass flow as an axial design, which is why we mostly see radial compressors in smaller engines, like those in helicopters and turboprops (i.e. small power, hence small mass flow).

  2. Thanks for this article series.
    It combines general explanation with state-of-the-art numbers, and is very helpful in understanding future developments.
    Although many of these numbers are probably public, most of us will probably not find them as we don’t look for them.

    Is it possible to present some “historic” numbers from early large bypass aircraft? I think of engines such as the JT9D or CF6-50. For me as an engine layman it would be interesting to know which areas have been improved and further can be improved in future engines and which technologies are required for that.

    • Hi Schorsch,

      good idea. I will do that in the next article. As we are still going to talk about compressors then we won’t miss anything. I will give data for a number a popular engines as examples.

  3. What is the pressure gradient of incoming air for large turbofan engines forward (several metres) of the intake when operating near maximum revs?
    It is probable that this pressure may be less than ambient pressure, and may be lowest in tropical latitudes around midday.

    For power plants mounted under the wing, this lower pressure may rob the wing of adequate lift to prevent stall conditions, requiring pilots to seek a nose down attitude to increase airflow over the wings.
    This could be hazardous at takeoff.

    Power plants mounted above the wing may improve lift if mounted in the optimum position, after research using the Navier-Stokes equations to look at the pressure gradients.
    A second foil (wing) may be placed above the engines to increase wing area and provide additional lift for the increased engine power, enabling greater payload on existing airframes.

Leave a Reply

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