I'm studying the fundamentals of jet propulsion and its applications. I really have this next question and haven't found an answer yet: why do jets use a shaft (or more) to connect the turbine(s) to the compressor device(s), but not a "rotating drum" instead?

What I call a "rotating drum" (I sketched it below) is a thin, rotating cylinder that has built-in rotor blades at the beginning and exit.

I'd like to know if this is a feasible arrangement, and what would be the disadvantages of it. I think this might be unfeasible because:

  1. There would be a lot of thermal stress at the middle of the drum.

But if we were to overcome this drawback somehow, I think we would benefit in the following ways:

  1. Less weight, I think, with respect the traditional shaft system.
  2. More internal space (no hollow) for other demanding components, like the combustor. The combustor could be made a bit larger, thus increasing the burning efficiency (more reaction time for reactants, ...).


I don't know how correct/wrong I am. This is a question out of pure curiosity and I wanted to get some feedback on it. Should I go ahead and start working on a small scale prototype of it? Hope my question here is not too basic.

PS: The sketch I made is just that: a sketch. Therefore, important things like the relative number of stages compressor-turbine, or the increasing/decreasing areas for each of those two components, do not appear.

  • 1
    $\begingroup$ The drum would still need support and bearings between its rotating part and the support structure. The likely place for this is in the center where a shaft would go. $\endgroup$
    – Steve H
    May 7, 2016 at 17:20
  • $\begingroup$ Either that, or a "guide" in the center, that impeded the horizontal move and where bearings would be placed. $\endgroup$ May 7, 2016 at 17:37
  • 3
    $\begingroup$ If I understand your drawing correctly, the "turbine" and "compressor" portions are fixed and the "rotating cylinder" portion spins around it, and the pieces labeled "S" would be the stators or guide vanes? In that case, the bearing surfaces would be the tops of the stators and the "lands" between the rotor blades. I can see basic value in this for a single shaft engine, but it seems that multi-shaft engines are the future. Interesting concept, nice drawing, and excellent hand writing! $\endgroup$
    – FreeMan
    May 7, 2016 at 17:49
  • $\begingroup$ Thanks FreeMan, too kind of you! For multi-shaft engines we could basically think of using coaxial rotating drums (but I'm not sure of that). Your understanding of the sketch was very precise. $\endgroup$ May 7, 2016 at 18:05
  • $\begingroup$ Centripetal force will tear the drum apart. $\endgroup$ Dec 9, 2018 at 15:44

3 Answers 3


There are a few very practical reasons why turbine engines use inner shafts/spools.

First, jet engine spools are very high speed rotating parts, this means they are subjected to high amounts of centripetal force. Centripetal force is calculated by $F=mr\omega ^2$ were F is Force, m is mass, r is radius of gyration and $\omega$ is angular velocity. So obviously the larger the radius of gyration the higher the forces that the rotors must carry. To reduce the radius of gyration manufacturers try to push the mass of the rotors as close to the centerline of the engine as possible. In the picture below (source) I have circled the HPC, HPT and LPT rotors. You can see how they are very thin under the blades and then have a large bulb very close to the engine centerline. Moving all of this mass towards the centerline, lowers the radius of gyration, and thus reduces the centripetal force. By reducing the centripetal force, you can reduce the strength of the rotors as well, since they don't have to carry as much load. This allows you to remove mass from the rotors, further reducing the centripetal force and overall weight of the engine.


Next, you want the bearings to be as small as possible. This is because as the radius of a bearing increases the linear velocity of it also increases, via this equation: $v=\omega r$. The faster the linear velocity the higher the wear, the greater the friction and greater the heat generation. So there is a push to make the bearings as small as possible given other constraints. In the image below I have circled the bearings, the red circles are on the low pressure spool bearings, while the blue circles are the high pressure spool bearings. This isn't always done, but it is becoming very common for the aft high pressure bearing to actual ride on the low pressure shaft. Typically these shafts spin in the same direction, so the bearing speed is reduced by this equation $\omega_{AftHPbearing}=\omega_{HP}-\omega_{LP}$. Bearing friction reduces the efficiency of the engine, and bearing wear is a significant maintenance driver.


(This is getting long so I am going to truncate the rest) There would be many other issues as well, such of blade mounting (dovetails in compression instead of tension), blade design (compression instead of tension), possible increase in leakage paths, containment, stator structure design (now all of your stators are attached to a stationary shaft that can only be supported at the ends), control of variable stator vanes, getting the fuel into the combustor, how do mount the engine, etc.

Edit: Just reread the question and realized you were also talking about making the combustor bigger. Modern combustors are getting smaller and are much smaller than they were in the original jet engines. You can also see in the pictures above, they already don't use all of the space available to them and if they needed more space the arm (shaft) connecting the HPT and HPC could be lowered some more.


A rotating drum will have a higher inertia. When the pilot commands a thrust increase, spooling up the drum will take longer than spooling up a shaft. Also, a big drum is harder to balance well than a thin shaft.

However, what you propose is not so far from general practice: The high pressure spool on modern engines is quite big already, but it is still on the inside of the blades. See the cut-away picture of the General Electric Passport engine for example (source):

General Electric Passport engine

Here it is interesting to see that the core components are quite a bit smaller than the channel for the bypass air, but the high pressure spool uses all the space left by the combustor. You can also see that the low pressure spool must be thin to allow enough cross section for the first stages of the high pressure compressor. Using a drum here would not only increase the inertia of the low pressure spool, but also of the whole high pressure compressor. Putting the drum on the outside of the blades would put it in the way of bleed air lines, accessories driveshafts and fuel lines. Also, making the stator vanes moveable would be much harder. The cut-away should make it obvious that the diameter of the high pressure section was kept as small as practical.

  • 1
    $\begingroup$ The OD of the spools where the arms connect the individual rotors is determined by getting the arms in the correct position to seal the stator bypass. The higher the OD on the arms the thinner you can make them, as well, since $\tau =Fr$ so as the radius increase, the force goes down, which is why the arm is so high under the combustor. $\endgroup$
    – OSUZorba
    May 8, 2016 at 15:05
  • $\begingroup$ @OSUZorba: I'm glad we have an engine expert around here! $\endgroup$ May 8, 2016 at 15:40

You seem to be describing an Exoskeletal Turbine design. I'm no expert, only vaguely associated with the industry. They're attractive in part because your rotor materials only need to be strong in compression, so you can use e.g. ceramics. Further, the hollow inner volume can be reused for an alternative flow-path or engine type, which opens up options for sc/ramjet options with a conventional turbofan wrapped around it. NASA's investigated it, with major findings that the bearings technology didn't exist at the time and that support equipment would likely eat up any advantage gained, for roughly equivalent performance. Materials improvements may shift the balance either way.

  • 1
    $\begingroup$ Welcome to Aviation. Great information in your first answer! $\endgroup$
    – FreeMan
    Apr 21, 2017 at 18:52

You must log in to answer this question.

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