[Canberrauav] X8 catapult
christopher.d.gough at gmail.com
Thu Nov 8 10:04:42 EST 2012
Last night I did some reading on RCGroups about catapult launching
small "jets" (models with high wing loading, e.g. stall speeds
>20m/s). I learned 2 things:
1) heuristic: ratio of bungee force to aircraft weight should be ~5 (5G launch).
2) if you measure the angle of the aircraft while hanging on a string
by it's launch hook, and rotate it by 90 degrees, that will be the AoA
on launch (given a flat bungee). i.e. momentum beats aerodynamics
For our purposes, I think we should aim to have the vehicle exit the
catapult at "maneuver speed" - allowing the autopilot/pilot full
control as soon as it is in the air. I dont' know what that is for the
X8, maybe 17 m/s? By maneuver speed, I mean the speed that the
airframe can do any maneuver it's capable of (roll, loop, etc),
probably similar to it's cruse speed (not close to stall speed).
That's not absolutely essential, but desirable.
Assuming we can engineer constant acceleration, and we know our
maximum acceleration (5G? 6G?), then we can deduce our minimum rail
WRT force, bungee is very convenient, compact and accessory-free.
Those benefits might overrule constant acceleration. I also like
Matt's pneumatic suggestion, especially with a short, fully contained
piston combined with a purchase mechanism so everythings' boxed in.
Gliders/parragliders are sometimes launched with a slip-clutch/winch
device attached to a car. i.e. the car drives forward at no less than
desired launch speed, and the winch let's out line (not pulls it in)
as required to sustain the desired force. Recreational fish killers
use similar devices to allow them to use thinner, more surreptitious
lines. We could do something similar too. For example, if the line
went from the plane, through a block and tacke, to a fishing reel. And
a short fat bundle of bungees pulled the block, then the fishing reel
would put a ceiling on the force applied to the plane. If the bungee
only delivered slightly more force than desired, the slipping reel
wouldn't need to compromise rail-length. If the reel had an on/off
switch for it's slip-clutch, it could also be used to wind back the
On Thu, Nov 8, 2012 at 8:57 AM, Matthew Ridley <dr.matt.ridley at gmail.com> wrote:
> I gave this some thought last night. Some other options to provide the
> necessary force in 2-3m might be:
> - Jack mentioned the "compound bow" approach. How about a leaf spring
> from a old car, and form a sort of giant cross bow ? mount the track
> perpendicular to this. Pulleys and cams could be added to make the
> force more constant and increase the travel to 2m. This would be cheap
> to build.
> - Pneumatic ram spreading a block and tackle ? Should be easy enough
> to get 100-200 psi into a small reservoir and drive a ram about the
> size of a bicycle pump at 50-70kg over 50cm. If the reservoir is large
> enough, the force will be more constant than a bungee. With losses on
> a 1:4 block and tackle I'm guessing maybe 10-15kg of force over 2m ?
> - I was also thinking of a flywheel/cam arrangement that can be spun
> up to provide the energy. Where the excess energy goes is a bit of an
> issue... compressed air can hiss out a hole, bungee cords can flap
> about without too much harm.
> - Could we spread a block and tackle with a car starter motor driving
> a worm screw ?
> On 8 November 2012 07:54, Andrew Tridgell <tridge at samba.org> wrote:
>> Hi Chris,
>>> Ok, I'd imagined more of a "high start" style, like a glider (maximise
>>> range). Thats why i thought your hook was to far forwards, the gradual
>>> acceleration was to flat. If you just want to slingshot it where its
>>> pointing that's different.
>> yep, I was always hoping we'd end up with a catapult with a fairly flat
>> trajectory (say 20 degrees or so). I think that will provide us with a
>> launcher that can work in a small area, with it getting on the motor and
>> under pilot control as soon after takeoff as possible.
>>> Are you sure the attitude solution will be OK with massive
>>> acceleration? If the horizontal accelerometer saturates, "down" swings
>>> forward, "forwards" meets ground...
>> It should handle it, although I may end up adding a special case like
>> Curtis does and disable the accel based correction for the first couple
>> of seconds after launch, operating on gyros only.
>> The accels are currently set to +/- 8g full scale. We hit around 1.8g on
>> takeoff on Sunday with Stans bungee. Using your bungee we got to around
>> On the first launch we got to around 13m/s in 1 second before starting
>> the motor. On the second launch I started the motor at close to the 2
>> second mark, at an airspeed of 10m/s.
>> I'm guessing that with a catapult the acceleration would be much higher,
>> perhaps 3g. If that is the case, and we have a 2m ramp, then we would
>> get to 10.8m/s at the end of the ramp, after 0.37 seconds.
>> If we get 4g, then with a 2m ramp we'd get to 12.5m/s after 0.32
>> That assumes a constant force while on the ramp, which of course it
>> won't be. So we'd actually need a bit more force at the start.
>> If we measure the force on the bungee with a luggage scale, then 4g
>> accel on a 4kg model will show 16kg "weight" on the luggage scale.
>> I think if we use shock cord (I have 10m of 8mm shock cord), then if we
>> have it loop back once on a 2m ramp, and have it stretched so that it is
>> still under some tension when it has released all the way forward then I
>> think it should work out OK. I tested this by stretching my shock cord
>> from a pole with my luggage scales and it looks reasonable. We could
>> measure the force at different stages of release by pulling againt a
>> pole to see how much the force drops as it gets released.
>> Cheers, Tridge
>> Canberrauav mailing list
>> Canberrauav at canberrauav.com
More information about the Canberrauav