I was reading about how the F-35 is having trouble with the "fueldraulic" system. This made me wonder:

Why don't aircraft use pneumatic systems instead of hydraulic/fueldraulic systems?

There are several advantages of pneumatic (compressed air) over hydraulic (pressurised liquid) or fueldraulic (pressurised fuel) systems:

  • No need to carry hydraulic liquid. That saves weight and maintenance cost.
  • Half as much piping, as there is no need to return hydraulic liquid to the pump - compressed air can be released after use. That saves more weight.
  • Leaks are less dangerous. There is no danger of running out of hydraulic liquid or of leaking fueldraulic liquid catching fire.

There are disadvantages too: less pressure -- which means less power per unit of gas/liquid -- and less precision, as gas is compressible. But I feel like these problems should be solvable in a modern computer-controlled aircraft. So what's the problem?

  • 11
    $\begingroup$ I guess you nailed it in your question. Hydraulics operate at 3000+ PSI so small tubes can be used. You would need tubes at a lot higher pressure to do the required work that you will have to continuously replenish by bleed air. Furthermore, there is huge lags everywhere due to the low speed of sound in air, and no matter how modern your control systems, there's a limit on how you can control things, called the 'waterbed effect'. $\endgroup$ – Sanchises Sep 30 '15 at 17:49
  • 6
    $\begingroup$ Aircraft do use pneumatic systems to power engine starts, the air cycle machine, wing anti-ice, pressurization, etc. That it also has a hydraulic system should shed some light that it is needed since the pneumatic stuff is already there. $\endgroup$ – casey Sep 30 '15 at 18:58
  • 1
    $\begingroup$ There are pneumatically actuated systems on at least some aircraft -- early 747s use pneumatic actuators for the LEDs, and some smaller aircraft have pneumatic L/Gs. $\endgroup$ – UnrecognizedFallingObject Sep 30 '15 at 22:27

The big disadvantage here is the loss of precision due to the high compressibility of gas compared to liquid. Because gases are highly compressible, they provide a buffer to changes in pressure commanded by the operator to move the piston in the cylinder. That poses two problems; first, it means that the pneumatic cylinder doesn't respond instantly to pressure differentials, because the differential must first overcome the cylinder gasket's static friction. Second, it means that the movement of the cylinder is more easily opposed as long as whatever force opposes the gas pressure can overcome said pressure without causing whatever the pneumatic system is controlling to fail.

To overcome these shortcomings, most pneumatic systems run at very high pressures, so that the pressure differential between the two halves of the cylinder readily overcomes static friction and any other opposing forces. However, that creates another precision problem; high-pressure pneumatic cylinders are essentially two-state systems; the piston or actuator is typically at one or the other of its extremes of movement, and transitions between them very quickly as gas pressure is applied to one side or other of the cylinder.

None of these behaviors are desirable for aircraft controls; instructors labor daily to teach their students not to ham-fist the controls, instead using a bit of finesse to get the plane to do what they want in a smooth, controlled fashion. Why then, would you want to undo all that finesse with a control system that can only move the surface to the extremes of its travel?

Hydraulics, by contrast, allow a much higher degree of finesse. Because liquids don't readily change density, the pressure changes within a hydraulic cylinder require much more force to oppose, but by the same token, as the volume changes the pressure on the side being supplied with fluid decreases rapidly. This allows a hydraulic cylinder to be positioned much more accurately, regardless of any external forces acting on the system. The disadvantage is hauling a fairly heavy liquid up into the air, and having only limited capacity to replace it if any of it leaks.

Electrical actuators are a common solution to that disadvantage, especiually in light aircraft. Electrical actuators use an electric motor or servo to provide the mechanical action. These actuators can be controlled with a high degree of precision, and their "supply system" is just an electrical circuit, no heavy and complex hydraulic lines and cylinders. Their disadvantages are a tradeoff between speed of movement and maximum applied force while moving; you can either make an actuator that moves very quickly, or an actuator that will move no matter how much force is opposing the movement, but you really can't do both. They're still useful in light aircraft to control flaps (with a cable system used for the main surfaces), because they allow for precise amounts of extension or retraction, and don't have to instantly respond to input like the primary control surfaces do.

There is something on the horizon that could make pneumatics feasible for aircraft. Hydraulics systems were recently improved with the development of the electrohydraulic servo valve. This system uses a variable electrical potential (voltage) to move a hydraulic cylinder by a prescribed amount proportional to the voltage applied. Pure electrical servos have been around for decades, but the maximum amount of force available from a servo is insufficient for large airliners, while for smaller aircraft the servo motor's relatively high weight compared to simple cable controls is a disadvantage. The electrohydraulic servovalve concept is used in newer large aircraft to replace pure hydraulic or cable/hydraulic hybrid control systems, because the hydraulic system can now be controlled by an electrical circuit instead of hydraulic lines or tensioned cables coupled to the control column. This allows for "fly-by-wire" aircraft such as most Airbus airliners as well as most fighter jet designs of the last 40 years.

A similar concept is under development for pneumatics, allowing the precise placement of an actuator using pressurized gas in response to an electrical voltage. This would provide all the advantages of an electrohydraulic system, with considerably lighter weight and faster response, but still having the disadvantage that a significant opposing force could prevent movement of the actuator especially as it approaches the desired position. Whether that will be an issue in a large aircraft remains to be seen, and the weight savings of losing the hydraulic fluid might not be worth it, but if the tradeoff is acceptable, it would further increase range or payload of the next generation of passenger aircraft, with the added safety/reliability feature of being able to compensate for a slow leak in a pneumatic system by simply adding more air with a compressor pump.

  • 3
    $\begingroup$ +1 for having a detailed answer instead of lazily cramming it in a comment. Guilty as charged... $\endgroup$ – Sanchises Oct 1 '15 at 8:28
  • 2
    $\begingroup$ A pneumatic cylinder could probably be better than a hydraulic one in cases where the objective was to modulate force rather than position; if one were to design the control linkage for a control surface so that the force generated by the surface would be some multiple of an applied force, using pneumatics to apply that force would mean that if turbulence caused the force produced by a certain angle to change, the surface would move in response without the change having to propagate through the plane's control system. In theory, nicer than hydraulics, but oscillations... $\endgroup$ – supercat Oct 1 '15 at 17:31
  • $\begingroup$ ...are probably far harder to control than with hydraulics (if hydraulics move a control surface to a particular position, having it stay precisely where it is would cause the effects of turbulence to buffet the aircraft, but if it's not moving it can't oscillate). $\endgroup$ – supercat Oct 1 '15 at 17:33
  • $\begingroup$ @supercat exactly what I was thinking as a counterexample. But in many applications oscillations like that would cause system failure. There are too many disturbances occurring during the time the system is correcting for these oscillations. Too many calculations and reactive inputs, never enough time or stability. $\endgroup$ – BAR Oct 2 '15 at 1:17
  • $\begingroup$ @BAR: Automotive suspensions have traditionally used hydraulics to control oscillation, but some newer systems are using more active controls. I'm not sure to what extent such a thing could be useful on aircraft, or whether there's so much unavoidable coupling of turbulence to the fuselage by the main wing surface that having control surfaces move in response to changes in airflow would do little to improve the ride. $\endgroup$ – supercat Oct 2 '15 at 15:23

One of the first reasons that comes to mind is air volume. Remember that a plane may be sitting on the ground on an 80 °F (27 °C) day, and take off and climb to 35,000 ft where temperatures of -50 °F (-46 °C) may be present. The air in the system would lose volume as it cooled off and would alter the position of the control surface (let's say flaps) with out any control input. Fluids are less susceptible to this problem. Granted this could be controlled, but it would still require a regulation system.

Leaks can also be easier to find in a hydraulic system since you can either

  1. see fluid leaking out
  2. put additives in that can be illuminated under certain lights

Pneumatic leaks are often found by rubbing soapy water on a joint and watching for bubbles (at least that's how I find them). Sometimes they can be hard to track if they are in awkward places.

  • $\begingroup$ The pressure differential is also an issue with the altitude change, though less of one considering the nominal operating pressures of pneumatic systems; at 50,000 feet, the pressure differential between "pressurized" and "unpressurized" sides of a pneumatic actuator is 14 PSI higher than at sea level. Now, if the system operates at 100 PSI anyway, this is fairly trivial. $\endgroup$ – KeithS Sep 30 '15 at 18:41
  • $\begingroup$ There are ultrasonic devices for finding leaks in pneumatic systems. They are fairly expensive though. And of course the non-leak-related reasons are far more significant. $\endgroup$ – Doug McClean Oct 1 '15 at 2:01

Yes of course (I am user12000 by the way :D) pneumatics are fast, cheap, and light but do not have good precision and you have to carry pressurized tanks called reservoirs (that means you need space) and you need to fill your reservoir (that means you need a compressor it means space again). When you compress air it heats up (it means a cooler system which means space again). You can use it again if you don't use it frequently and you want cheap and light. Hydraulics are heavy, expensive, can use high force, and have good precision. You don't need to use a compressor because its uses fluid it needs a pump. Pumps are to smaller than compressors and they produce less heat. You also need a small reservoir (reservoir is needed hydraulics because of protect the system from stress by expansion of fluid because of heat up of system or vice versa).


Yeah, precision would be the deal breaker. This is why excavators, for instance, use hydraulics because you can deliver a lot of power with very minute movements. Good operators could pick up a quarter then knock over a tree. So while in flight, control surfaces are subject to enormous pounds of air pressure but must move only a few inches and with a within a rapid response time. A liquid is going to have the viscosity for more "fluid" movements which is more susceptible to instantaneous responses. You'll see pneumatic systems a lot in plants for valve applications which will be fully open or fully closed. Because air is so light is just can't deliver the same power and precision.

Besides fly by wire makes any of this talk rather obsolete. You can get all the electrical power you need to extremely high HP/torque actuators that will deliver just as much force with even more precision and greater response time.

I'm not sure but I wouldn't doubt the wheel braking system might be pneumatic like over-the-road 18 wheelers. Otherwise the only pressurized air you're ever going to see on a airliner will be cabin pressure and oxygen delivery.

  • 1
    $\begingroup$ Viscosity has nothing to do with this. Liquid is going to have higher viscosity, which means more power will be lost to internal friction as the fluid flows through the pipes. The reason for using liquids is their incompressibility, which allows them to transfer large forces (pressure) with minuscule changes in position. $\endgroup$ – Jan Hudec Oct 2 '15 at 5:04
  • 1
    $\begingroup$ Fly-by-wire doesn't (currently) uses electrical actuators, but still hydraulics, or a mix of hydraulics and electrical for redundancy. When the system falls back on electrical, the control may be limited or slow. $\endgroup$ – mins Oct 2 '15 at 5:09
  • 2
    $\begingroup$ Fly by wire aircraft still use hydraulics for most of their actuators, because electric motors can have high torque or high precision, but they don't have both. $\endgroup$ – Jan Hudec Oct 2 '15 at 5:11
  • $\begingroup$ @JanHudec Why not? $\endgroup$ – Koyovis Feb 16 '18 at 23:45
  • $\begingroup$ @JanHudec The thing is, if looking at energy consumption, a hydraulic motion system for a simulator always consumes maximum power, whether the payload is moving or not. The excess energy not used for motion is transformed into heat at the servo valves. An electric motion system only consumes the energy directly used for motion (or compensating for gravity if no air spring is present). The force feedback systems at Level D flight controls have been electric torque motors as well for the last 20 years: silent, clean, frugal with energy. Granted the motors are bigger than hydraulic actuators. $\endgroup$ – Koyovis Feb 19 '18 at 8:45

espacially harnesses include mechanical, hydraulics, pneumatics and electromechanical parts. Generaly Aileron, elevator, and flaps are controled by hydraulic servos because hydraulics have advantages precision and powerloss, but on landing gear doors are controled by pnuematics(not servos)because you don't need any position feedback you only want to know doors are open or not.

  • 1
    $\begingroup$ You bring up some good points, it would be helpful to include some additional explanations of the advantages and disadvantages you mention. $\endgroup$ – fooot Oct 26 '15 at 21:09

This brake discussion is old, but I don't see any answer pointing out that the 787 has electric brakes. (But see 5th paragraph of answer by KeithS regarding small planes.) From the 787 FCOM section 15:

Electric Brake System

The brake system is powered by four electric brake power supply units. The brake pedals provide independent control of the left and right brakes. Four Electric Brake Actuators (EBAs) are provided on each main landing gear wheel brake to control the application of braking force to the carbon disc. The EBAs are controlled by an Electric Brake Actuator Controller (EBAC). There are four EBACs that control all eight main wheel brakes, each EBAC controlling the brake force of a fore-aft wheel pair.

Source: Document Number D615Z003-TBC October 31, 2007 Revision Number: 4 Revision Date: February 15, 2010

Of course, Boeing talks up the electrical systems on the 787. Here is what they say about the brakes:

One innovative application of the more-electric systems architecture on the 787 is the move from hydraulically actuated brakes to electric. Electric brakes significantly reduce the mechanical complexity of the braking system and eliminate the potential for delays associated with leaking brake hydraulic fluid, leaking valves, and other hydraulic failures. Because its electric brake systems are modular (four independent brake actuators per wheel), the 787 will be able to dispatch with one electric brake actuator (EBA) inoperative per wheel and will have significantly reduced performance penalties compared with dispatch of a hydraulic brake system with a failure present. The EBA is line-replaceable enabling in-situ maintenance of the brakes.

In general, electric systems are much easier to monitor for health and system status than hydraulic or pneumatic systems; the brakes take full advantage of this. Continuous onboard monitoring of the brakes provides airlines with a number of advantages, such as:

Fault detection and isolation Electrical monitoring of brake wear Ability to eliminate scheduled visual brake wear inspections Extended parking times Because the 787 brakes can monitor the braking force applied even while parked, the electric brakes enable extended parking brake times by monitoring and automatically adjusting its parking brakes as the brakes cool.

How do they work? Here is a clue from PPrune:

The motors supply torque through gear assemblies to the actuators, which are rams driven by a jack screw. There is a latching mechanism to limit over-rotation and back driving, minimizing the current requirements for the EBS actuator motors.


Yes what's the problem indeed. It has been done, without any problems, with all the advantages you mention and without your cited disadvantage of low pressure. This article discusses the fully pneumatic design of the F27 and F227. Air at 3,350 PSI has some mighty fast action, plus it has stored actuation power which a hydraulic system does not have. Hydraulics can deliver a high pressure at a low rate: the pump rate. Hydraulic accumulators store a bit of extra pressurised oil so the system can very briefly exceed the pump rate - only briefly though, and the accumulator is also required to dampen out ripple. High pressure air can be delivered at a huge rate, for a much longer time.

The F27 has manually driven flying controls: an aircraft this size can be fully controlled using these. For larger aircraft, the force required to deflect the control surface can be generated by an actuator working at 228 bar, whether hydraulically or pneumatically operated.

I've spoken with some of the designers of pneumatic systems when I worked at the factory that made these planes. The only real difficulty they experienced during the design and implementation phase was the design of the controllers, a servo valve for a hydraulic system gives less headache than one for a pneumatic system. Just an additional engineering problem to be solved with an appropriate feedback loop.

And now we can design for the least problematic controllers of all, for electric motors & drives.


Not the answer you're looking for? Browse other questions tagged or ask your own question.