Why are winches not used for towing commercial planes up?

In this video a glider is pulled up by a winch on the ground, like this:

(Source)

I would say you could apply the same principles with bigger planes, possibly even using sustainable energy sources.

The plane would not use its engines at all to get to cruising attitude (unlike a suggested catapult in a previous post), so this part of the flight would be "for free". Using a back of the envelope type of calculation with 5% efficiency it takes 17000 litres of kerosine to lift a 747 to 10km (which agrees with this paper).

However, this would however cost at least (calculated below) 25km of towing line(s). If we use a UHMWPE fiber a 400g/m and 1.2 cm radius cable would be able to support 2x the 400kN thrust maximum output of a typical airliner.

Sure there are a lot of hurdles to overcome, but since some airports in cities cannot expand because of noise problems this seems like one solution.

Logistically this seems like a large extra challenge for the crew on the ground, but would this be possible?

• Comments are not for extended discussion; this conversation has been moved to chat.
– Farhan
Jan 16 '18 at 18:44
• You mostly answered your own question in your bounty post :: I'm still not convinced this is impossible, while not useful for less densely populated areas this could be useful for densely populated countries :: It's not impossible; anything can be done if someone has the will. But cable-launching heavy aircraft has very little usefulness, no matter the population. Why do you think it could be useful in densely populated areas? Jan 25 '18 at 22:32
• "I'm still not convinced this is impossible." Then it's an engineering question. How thick of a steel cable would we need to slingshot a 747 to cruising altitude? According to one of the answers (uncited! @mins), it's : too big. Jan 27 '18 at 1:38
• We calculated in the comments it could be done with a UHMWPE wire, with a factor 2 safety ratio the wire would have to be 1.2cm radius and 400g/m Jan 27 '18 at 13:39
• Gotta say, *boggle* at the bounty. One of the answers is three screenfuls of my browser; another is four. Both posted long before the bounty. How much detail do you want? Jan 29 '18 at 22:57

If you ever get the chance, do try a winch start in a glider. It is, to say the least, an interesting experience. I do not believe it would be accepted by the general public.

So, that is one more reason, it would scare the paying passengers.

• I have not tried it myself... but surely, in a glider it would feel a lot rougher than in a jetplane due to the lower mass? Jan 14 '18 at 21:14
• @leftaroundabout But surely, in a jet it would feel a lot rougher than in a glider due to the required higher acceleration. Jet has a much higher desired speed and you have a limited distance, so you will need a lot more "g". Mass doesn't even appear in this equation. Jan 14 '18 at 22:37
• +1 It looks pretty scary to me. Jan 14 '18 at 22:48
• @kubanczyk that's not the point. A glider can accelerated so quickly (and in particularly, so roughly, which I presume refers to a pretty sudden “kick in the back” at the beginning), because it's so light. That's good of course because it allows very quick takeoff and therefore much height to be gained without excessively long rope. For a fully fueled jet though, that much acceleration would simply not be feasible, you'd have to make do with less g spread over a longer way. Then again, that would make the rope length totally impractical, which is really the main reason this isn't done. Jan 14 '18 at 23:34
• I'd say you'd need about the same acceleration as you get now. More is better (within reason) because of the higher altitude you can reach with a glider and because you want the glider to fly and be controllable as soon as possible. Jan 15 '18 at 5:57

I won't prove it's not possible, but I'd would be surprised it is. I'll develop the reasons why and provide the physical framework you can use to compute a solution to have, with some effort, a final answer.

The glider case cannot be scaled up to an airliner

It works with a glider for two reasons:

• The height to be reached before the glider is able to gain altitude by itself is limited: 200 m or so. To be compared with the 10 km the airliners fly (as you want to reach the cruise altitude without the engines).

• A glider has a lift-to-drag ratio of 30, 60 for best gliders, a B747 has a L/D ratio of 17.

These two differences have huge interlocked consequences:

• When you raise a cable, the top segment of this cable must be able to resist the whole weight of the cable, plus the drag created, plus the force that the cable needs to transmit in order to move the aircraft forward. Each newton or kg added to the wire increases the required section, hence the weight, hence the section, etc.

• If the L/D ratio is lower, then the aircraft will create more drag when it will create lift to lift the cable, hence the force created by the winch must be larger to move the aircraft forward, hence the section of the cable must be larger, hence the first problem is aggravated.

Description of the problem

The cable curve will be a catenary, actual calculations by an engineer should prove the solution isn't possible with the material we have today (including a good candidate, Dyneema used by tug boats).

Kite analogy

Those interested may also try the Kite Modeler Java applet from Nasa, as in this configuration the airplane is mostly a kite:

• How to manage a safe corridor long of something like tens of km?
• How to reset the winch for the next launch?
• Effect of the wind on the cable, and corresponding jolts on the aircraft.
• Cable elasticity effect on the aircraft.
• How to continue operations when static electricity starts to accumulate in air, and you don't want to transform the system into a lightning rod?
• "L/D ratio of 17 cannot climb efficiently more than 1 km while moving forward by 17 km." That ratio does not tell you anything about the climb rate. Look at the 15% used here currently which is a lot more than 1km/17km Jan 14 '18 at 16:28
• Static isn't a problem: just make sure the entire airplane-cable-winch system is electrically bonded and grounded, and any static will dissipate as fast as it builds up.
– Mark
Jan 15 '18 at 1:40
• The altitude you can reach this way is hardly limited. My personal best is a bit over 700m (on a relatively long field with a strong headwind). Also, I do not buy your argument that the rate of climb is limited by the L/D ratio. A plan can climb much steeper then its L/D ratio: a glider (L/D > 30) can climb to about 1/3rd to 1/2 the runway length, depending on wind and operator skill of the pilot and winch operator. Jan 15 '18 at 5:51
• Nice math, but wrong entry data. L/D is for plane as closed system, expending stored potential energy to compensate drag loses. Winch is adding energy to the system. For a glider - all of it. Jan 15 '18 at 9:08
• I don't see how the L/D of 17 converts to a 170 km horizontally for a 10 km cruise altitude. For higher L/D (e.g. 30 for a glider) would it mean that more horizontal distance is required? What if we have an extremely efficient aircraft with L/D of 10 000. Would a meter climb by cable towing than require 10 km of horizontal distance? That doesn't make sense. I'd expect it to climb steeper Jan 15 '18 at 9:09

Sure there are a lot of hurdles to overcome.

Plenty. I've said it before, engineering fixes problems, not create ones.

But all the hurdles really don't matter. Gliders don't use winches because of limited runway, they use it because they don't have engines, so the comparison stops there.

And thus, there is no issue with commercial aircraft (jets and turboprops), they have a lot of power and already operate from short runways.

Noise

A city in close proximity to a small airport will not be happy when a stupendously heavy cable is released from the launched aircraft and falls on it. If there is an area for the drop, then surely there's an area for a longer runway and an area for a lower climb power setting.

• definitely one of the hurdles, but I would say doable. I come to 5 kg / m of soft textile using 8000 kN of max thrust of an airliner and 14 g/m for my dyneema ready made 22kN climbing sling. With parachutes and these kind of things I would say that is reasonable Jan 14 '18 at 14:10
• Good point ymb! So, assuming we want to get to cruising altitude of 9KM, we would need a 153km long cable and a path that long to launch it considering a 747 goes 180 MPH on takeoff and it would be 1000/ft minute climb rate to get to 30K feet. So, the government goes out and uses imminent domain to buy people's property, they have a search crew on standby to go running around the countryside to find where the cable lands, unwinding it from trees and such. Or maybe there should be massive deforestation in the region to accommodate the cable falling. By that alone, this is not realistic. Jan 14 '18 at 16:02
• Winch tow is not exclusively used because of short runways, but also because it is much cheaper than aero-tow (and start frequency can be hogh, too). Jan 14 '18 at 19:03
• "I've said it before, engineering fixes problems, not create ones." As an engineer, hahahahahaha! Jan 15 '18 at 16:51
• @Ruud3.1415 You would need to use a more capable cable than that. While climbing slings are rated to 22kN, you typically have to replace them after a few brutal falls because you no longer want to trust your life to them. A cable like this would receive a lifetime of climbing harness abuse every day. Jan 16 '18 at 0:42

(I hesitated to write this answer because I couldn't find official sources, but in the end I decided to do it; the source is that I flew gliders, instructed, and dealt with their maintenance for many years)

Something not mentioned in the other answers is the stress on the airframe. A glider that is winched regularly has a reduced lifespan. Like I said above, I don't have concrete sources, but I do remember being involved in the extension of the end-of-life of a Blanik, and to determine how many hours were allowed; the manufacturer considered regular winch launches on the same level as regular aerobatics (that is, as something that substantially reduced the life span of the airframe).

• 'cause you'll break the plane +1 Jan 27 '18 at 1:51

In addition to the other excellent answers, two more strikes against cable launch in the commercial aviation industry:

1. Need control and power for emergencies
2. Additional point of failure and loss of runway
3. Runway damage

First, an aircraft will always have full power upon launch whether cable launched or not. This is to give the widest possible safety margin in case a problem comes up.

Second, a cable launch mechanism adds an additional point of failure. Not only could we end up not being able to launch due to a cable launch mechanism problem, but it could prevent that runway from being used for takeoff or landing depending on the problem.

Lastly, the cable and its attachments must be necessarily heavy since they will be slamming into the ground or perhaps a plate designed to absorb such impact when released. However not all releases will be ideal, and if the cable or its attachment hits the runway it will be compromised, potentially halting all traffic on that runway for many hours.

All these things can be handled, but are additional factors that would prevent an airline and airport from adopting such a system.

However, there's an additional positive offset:

Takeoff uses a significant amount of fuel. If the takeoff can be energized from a ground power source, particularly where electricity is cheap, you may be able to offset some fuel and its associated costs for each flight, which could result in a significant savings over industry flights.

This would also reduce (or at least relocate) carbon emissions.

• an aircraft will always have full power upon launch whether cable launched or not - Not quite accurate, especially in commercial aircraft. Takeoffs are quite often done at less than full power for various reasons. Jan 25 '18 at 22:17

As a glider pilot who mainly uses winch launching and also a professional engineer I can see both sides of the issue.

You can solve the launch issues by throwing engineers and money at the problem but there is a glaring issue that does not seem to have been covered: cable breaks.

Despite having over engineered cables and linkages, cable breaks do occur, I have had a couple and they can be sphincter-clenching moments but we train for them and we survive. We survive them mostly because we are flying gliders, aircraft that are specifically designed to be flown without power.

A break at low level and I just land beyond the winch, this would need a massively long runway with the winch partway up, a mid level break and I turn and do a cross-field landing which is easy as my airfield is very wide but a commercial airport has fairly narrow runways. A higher level break and do a quick circuit of the field and land normally. An airliner is not a glider and so the land ahead is possible but the circuit and land normally would not be possible due to the glide characteristics of an airliner. A cross field landing would require a square runway which is massively expensive.

...and then you have the problem after a successful cable release of several tons of cable dropping to the ground. In gliding this is done by means of a parachute under tension from the winch. In the event of a winch failure at cable release then you need to have a clear area all around the winch the same diameter as the release height in case the cable drop onto something or someone. Unlike gliding this cable will be very very heavy.

• any idea how much time it would take to spin up the engines as a backup-strategy? Jan 29 '18 at 13:04
• About the falling wire when it would break, I calculated that the terminal velocity of the 1.2cm radius 400g/m wire I mentioned above (2x strength of full thrust engines) would be around 86km/h or 53 mph. I think getting hit by this would not be likely to seriously injure someone and the wire could easily be made even thicker to reduce terminal velocity. Jan 29 '18 at 13:26
• As a glider pilot engines are voodoo to me but I would guess that the engines would be running and ready but if you were low enough to the ground and had a break the time between full power and an uncontrolled stall would be uncomfortably close. Jan 29 '18 at 13:27
• @Ruud3.1415. That is most likely correct for a linear fall but there can be a 'bullwhip' effect when the wire drops and the end flips out at a very high speed. In addition you have a weight on the end for the attachment gear (assuming the wire falls intact) which can do a lot of damage. The wire I launch on for a 500kg glider is about 1cm diameter. I suspect that for safety an airliner version would be grossly over specified and likely to have protective sleeving to prevent the abrasion from being hauled along the runway at speed. I take off on a grass field and the wire still gets worn. Jan 29 '18 at 13:37
• Just as an extra, gliders do not fly as they are launched, they 'kite'. On a cable break your glider instantly stalls and you have to take a (very long) few seconds letting the nose drop and then dive to get airflow over the wings so that the control surfaces start to work. The flight characteristics of the glider allows for a quick orientation change and control recovery. A large aircraft would respond much slower and a low level break would not allow enough time for it to re-orient and land even with engines spinning up to speed. Jan 29 '18 at 13:43

Cables are heavy, let's do railguns

Essentially gliders are light and only need to be pulled to a modest height, an airliner is vast and you could only hope to save a fraction of the takeoff energy. The plane couldn't be made smaller unless it only took off from such places. And the cable to pull an airliner would be incredibly heavy, and you'd only get 3km or so (the length of the runway) to use unless you think airports always have space beyond the runways.

A railgun, though, would be more fun.

How fast can we go in 3km?

Imagine we accelerate at 1g (total force on the passengers is then 1.4g) for 3km (reasonable international airport runway). v^2=u^2+2as -> final v=240 m/s or 540 mph which is pretty close to cruising speed. So we don't need even a 3km runway for this, or we could accelerate more gently.

There is a hitch - we're still on the ground. So in fact the best thing is to assist the aircraft to a takeoff velocity (v2) and let the rest of the takeoff proceed as normal.

If that's so clever, why isn't it being done?

It is in development. The EM Assisted Launch System for the US Navy, to replace catapult launch from carriers.

But, as this answer explains, the proportion of an airliner's energy used to take off is a small fraction of the total energy used, so you are likely to lose any benefits in additional weight from implementing such a thing.

• This idea is similar to this post. While it could reduce the length of the runway needed, the energy an noise savings would not be fairly significant. Jan 16 '18 at 15:06
• Good point, have updated. Jan 16 '18 at 16:21
• I don't think your answered the question which includes "The plane would not use its engines at all to get to cruising attitude".
– mins
Jan 27 '18 at 22:21
• @mins: Well, the basic answer to the whole requirement is No, so it seemed better to elaborate on the approximate concept. Jan 29 '18 at 10:15

A fairly obvious point that seems to have been missed in the other answers: how fast can you cycle the winch? A busy airport might have planes departing every couple of minutes. To get to cruise altitude, your winch has to reel in a considerable length of cable. This then has to fall to the ground after the aircraft releases it, which takes a minute or two at best, more if it has a parachute. Then a ground crew has to grab the end of the cable, reel it back to the departure end of the runway (which requires a vehicle), and attach it to the next plane. Say you can do all that in a really optimistic 10 minutes: that means you've cut your airport's peak departure capacity to 20%.

Then you have the cost of all this infrastructure, plus you have to duplicate it so you can take off in either direction. And that's not even getting into what you might do for intersecting runways (or even parallel ones), or how the cables might affect landing traffic...

• Clearly you would have more than one winch-reel - deploy No.2 while you are launching No.1, recover No.1 while launching No.3 ... Oct 13 '20 at 12:16

I believe it is a good idea, and that it will happen eventually, but not anytime soon though. With the airplanes we have now it is simply not worth it economically. We say that bringing extra fuel on a plane is very expensive. It is expensive because, if you bring extra fuel to extend your range, you need to get it in the air, and then carry it the whole way until you need it. It is not the price of the fuel that is expensive, it is having to carry it until you need it that is expensive. The fuel for take-off and climbing is used right away, so it is not expensive at all. The electricity bill for the winch would probably cost almost as much as the fuel you are saving.

The system seems perfectly feasible though. If it works on a small scale, for fighter jets and gliders, then why would it not work on a larger scale? I believe air traffic will slowly start to move towards electrically powered planes within the next few decades. For electric planes, a ground assisted take-off would be much more beneficial, or even essential. A winch could take the plane to above the clouds, and there it can fly to it's destination on solar power. If it had to get there itself, it would need huge battery packs.

• while this would be ideal, using a 100% efficient solar panels providing the current ca. 5MW needed for an airplane during flight would need around 6000 squared meters of solar panel area. Jan 29 '18 at 15:02
• If we take the Shockley–Queisser limit you will need 3 times more than that. But, then again the 5MW would include takeoff Jan 29 '18 at 15:04
• @Ruud3.1415 There is still a lot of work to be done, but i think we'll get there: theatlantic.com/photo/2016/07/… Jan 29 '18 at 15:15
• "If it works on a small scale, for fighter jets and gliders, then why would it not work on a larger scale?": I can drop a 5 cm miniature car from 3 times it's length, it won't break. But I can't obviously do it with an actual car.
– mins
Jan 29 '18 at 17:39