Many jet aircraft travel at speed of about Mach .8 which at an altitude of 11 km represents an airspeed of about 300 kt or 150 m/s.

  1. If fuel was injected into air at this speed, the mix would just leave the engine before being burnt, combustion would occur behind the aircraft. From different readings, I believe the correct airspeed in the combustor must be around 10 or 20 m/s to maintain the combustion within the combustor.

  2. In the compressor section, air from the intake is forced to rush into a convergent annular tunnel to compress it. Intuitively, this design won't slow air, or not slow it enough.

    enter image description here
    General Electric J85-GE-17A, source: Wikipedia

How engineers are able to slow air without decompressing it?

  • $\begingroup$ Engines aren't drag-less, meaning the air doesn't pass through them unimpeded. The compressor section slows the air down quite a bit. In supersonic aircraft the intake design also plays a large role as the airflow has to be slowed to subsonic before it enters the compressor stage. $\endgroup$
    – Ron Beyer
    Nov 10, 2016 at 15:18

4 Answers 4


In general, aircraft jet engines have diffuser section(s), that reduce the velocity of the incoming air before it enters the combustion chamber.

Jet engine velocity

Sample velocity profile in jet engine; image from Fundamentals of Gas turbine engines

In some cases, the the diffusers are before compressors or in the stages itself- but the end effect is reduction in velocity. Even in this reduced speed, combustion is not feasible as the the speed of burning kerosene at normal fuel-air ratios still lower; hence any fuel lit even in the pre- diffused air stream also would be blown away.

In order to overcome this, a region of low axial velocity is created inside the combustion chamber using swirlers and recirculation. It helps that the fuel is burnt with only a portion of the air entering the combustion chamber.

Combustion chamber

Image from aeromodelbasic.blogspot.in

Basically, the airflow entering the chamber is divided into multiple parts, which enter the chamber at different times and places, so that the overall airflow has low axial velocity, while promoting recirculation. From Combustion process:

Approximately 20 per cent of the air mass flow is taken in by the snout or entry section. Immediately downstream of the snout are swirl vanes and a perforated flare, through which air passes into the primary combustion zone. The swirling air induces a flow upstream of the centre of the flame tube and promotes the desired recirculation. ...

Through the wall of the flame tube body, adjacent to the combustion zone, are a selected number of secondary holes through which a further 20 per cent of the main flow of air passes into the primary zone. The air from the swirl vanes and that from the secondary air holes interacts and creates a region of low velocity recirculation. This takes the form of a toroidal vortex, similar to a smoke ring, which has the effect of stabilizing and anchoring the flame.

  1. Combustors usually feature recirculation zones in which the combustion takes place either completely or which at least anchor the flame. This can be done by several designs
    • Sudden jump in cross-section in combination with vorticity
    • Flame-holder bodies
    • Central displacers
    • Vorticity brake-down
  2. The blade channel of a compressor is actually divergent. Deceleration of a flow goes along with an increase of the static pressure (see diffusors). Additionally, a compressor transfers work into the fluid, thus further increasing the fluid density. The divergent cross-section of the annular channel between hub and casing is designed to keep the axial velocity somehow level. High axial (or more correctly meridional) velocity levels cause higher friction losses. Low velocities reduce the transferable power of a compressor stage, thus requiring more stages.
  • 2
    $\begingroup$ While I agree with your answer and up voted it. It would be better to add some links to backup your information. $\endgroup$
    – Notts90
    Nov 10, 2016 at 11:24
  • $\begingroup$ "The blade channel of a compressor is actually divergent". This is something that in itself should be explained, as the channel is seemingly convergent before the diffuser, each stage of the compressor having a larger channel cross section forward than rearward. $\endgroup$
    – mins
    Nov 10, 2016 at 21:47
  • $\begingroup$ @mins: The annular space between hub and casing differs from the blade (or vane) channel that is defined as the passage between two blades (vanes) of the same stage of a compressor/turbine. You can estimate the flow cross section by zooming into the image attached to the answer and regard compressor/turbine at shaft height. You can see that the channel in a compressor is (slightly) divergent whereas the turbine channel is drastically convergent. $\endgroup$
    – Chris
    Nov 11, 2016 at 6:44

enter image description here

By SidewinderX (Own work) [CC BY-SA 3.0 or GFDL], via Wikimedia Commons


The swirler is a part of the combustor that the primary air flows through as it enters the combustion zone. Its role is to generate turbulence in the flow to rapidly mix the air with fuel.

Fuel injector

The fuel injector is responsible for introducing fuel to the combustion zone and, along with the swirler, is responsible for mixing the fuel and air. There are four primary types of fuel injectors; pressure-atomizing, air blast, vaporizing, and premix/prevaporizing injectors. Pressure atomizing fuel injectors rely on high fuel pressures (as much as 3,400 kilopascals—500 psi) to atomize1 the fuel.

1 While atomize has several definitions, in this context it means to form a fine spray. It is not meant to imply that the fuel is being broken down to its atomic components.


So, mainly these two components ensure that the fuel is thoroughly mixed with the air for the combustion without losing fuel to the downstream.

  • 1
    $\begingroup$ You have been the first to mention the swirlers, however you answer mostly focuses on their role in improving the mix, but you may develop their role in maintaining the recirculated air within the combustor. +1. $\endgroup$
    – mins
    Nov 10, 2016 at 19:51

How engineers are able to slow air without decompressing it

Flow speed is energy, so slowing down the flow will increase its pressure.

At subsonic speeds the first part happens already ahead of the intake when the airplane approaches. This pre-compression is very efficient because it happens in the free stream, and the engineers design the intake such that it will only swallow part of the air flowing towards it at cruise speed. Within the intake the flow decelerates further such that it will enter the first compressor stage at a flow speed of Mach 0.4 to 0.5. Typically, 98% of the flow's kinetic energy can be converted into pressure this way.

In the compressor section, air from the intake is forced to rush into a convergent annular tunnel to compress it. Intuitively, this design won't slow air, or not slow it enough.

Within the compressor the density of the compressed air rises together with the pressure. Therefore, the converging geometry within the compressor does not accelerate the flow - it is actually slowed down further - and only follows the changing volume. The density and volume ratio at isentropic compression are proportional to the pressure ratio (index 0 denoting the initial and index 1 denoting the final state) like this: $$\frac{p_1}{p_0} = \left(\frac{\rho_1}{\rho_0}\right)^\kappa = \left(\frac{V_0}{V_1}\right)^\kappa$$ with $\kappa$ being the ratio of specific heats (approximately 1.4 for air). The 8.3 compression ratio of the J-85 in your picture will compress the air to 22% of its initial volume.

The final step in the deceleration of the flow happens in the cross section leading from the compressor to the combustion chamber, which is called the diffusor. Here the cross section carefully widens to slow down the airflow without separation. Around the fuel injectors you will find the lowest gas speed in the whole engine. Please see this answer and this answer for more details - I am told not to copy relevant parts of older answers into new answers.

  • $\begingroup$ +1, but I'm confused why you voted to close this question as being a duplicate of another question you also answered, and still add a new answer here. Do you think your first answer is not good enough? In that case why not just update it instead of posting a new one? $\endgroup$
    – mins
    Nov 11, 2016 at 13:19
  • $\begingroup$ @mins: No, I felt that aeroalias' answer is incomplete and that this question needs an adequate answer. The answers to the other questions should be adequate, only your suspicion about the converging cross section along the compressor needed to be addressed here in detail. However, anyone else looking for flow speed information should be adequately served with the other answers, hence the close vote. $\endgroup$ Nov 12, 2016 at 2:16

Not the answer you're looking for? Browse other questions tagged .