Society of Robots - Robot Forum
Mechanics and Construction => Mechanics and Construction => Topic started by: MangoBot on July 25, 2009, 11:53:31 AM
-
so i wa thinking if it was possible to attAch a weight to a solenoid and attach it to a small hang glider, there would be sharp ir scanning for objects and two solenoids would push or pull weights to change direction(one for left and right, and one for higher and lower) i was thinking of dropping it off of a rc plane, and maybe have it trying to get to a place using a transmitter receiver.
-
using solenoids for weight shifting would be a good idea. you could also use an 3-axis accelerometer to keep it stable in flight (it might be best to use 2 or 3 of these to compare readings, especially with things like pitch and raw)
as for using the sharp IR, i dont think that would work, the sharp IR has a limited ranged (usually a meter) and that wouldn't give you hang glider enough time to respond, a hang glider would be traveling at about a meter per second.
For obstacle detection you would need something with a greater range
-
ok, would something like sonar work, and for the material i was thinking of using small plastics rods attached to garbage bags(cut up into sheets) the rods would be the frame,
-
I don't know if sonar has that kind of range either, but, i've never used sonar so, it could.
As for building a hang glider, I would make it like one would make a kite (i.e. use carbon fiber spars and ripstop nylon for the sail). You can search for kite building, i know there are various large communities related to kite building, and i've looked into it before, so i know you can find information on it.
-
yeah, i have 2.5 acres to experiment with
-
and you are very lucky to have 2.5 acres to test and experement on.I sure wish i had that much and , do you have farm or something.
-
no, but we do have a garden, I might even make a remote control glider to launch it, i might use two mercury tilt things
-
having it remote controlled would be awsome, And you are so lucky to have that much garden ;D
-
yeah having switches to adjust the position of the solenoids would be cool, a airplane type glider would be cool, actuall i may make a r/c plane instead of the glider and maybe slowly make it more autonomous
-
that sounds awsome ;D
-
yeah, i wonder what the first system should be?
-
well, before you even think about starting this... take into consideration weight of solenoids, batteries, control systems and sensors. this may not seem to weigh a lot, but when its on a glider... it will.
have you considered using flaps instead of weights?
-
yeah, that is what i will use for the plane
-
Definitely do lots of research before starting this project. Not only should you look into basic aeronautics (for flight control theory), I highly recommend looking at current R/C planes. I'm 98% certain I saw a remote controlled hang glider before, you could look at how they did it (or even buy it and modify it). If you already have an RC glider, replacing the RC bit w/ a microcontroller wouldn't make much of a weight difference.
-
smash has a good point, weight is going to be amassive issue if you build a glider.
And remember do the math to work out the glide ratio. to see if it is a viable glider
-
Instead of moving the weight why not twist the wing?
Also you will have to learn how to fly hobby-grade RC planes.. they are quite difficult so its best to start out with one that you won't get frustrated with. Check out the hobbyzone super cub.
-
Thinking about it some more, I would consider building more of a traditional glider with traditional control surfaces (rudder, elevator, etc). Something like that would be easier to control I'm thinking.
oh, and for testing considerations, your 2.5 acres probably wont be big enough, in order to have a plane (or glider) that could support the extra weight of a controller and sensors you're going to need a bigger plane, bigger planes require more space. I'd recommend finding a large open field to test in.
Also, you might want to consider a robot controller RC plane first, You could probably buy an RC plane that you could learn to fly on and then set it up to fly on its own maybe?
Also you will have to learn how to fly hobby-grade RC planes.. they are quite difficult so its best to start out with one that you won't get frustrated with. Check out the hobbyzone super cub.
I agree that you will definitely need to know how to fly an RC plane before you start this project, I'd recommend you towards the Firebird Commander, its not terribly expensive and has everything you need.
-
for the moving flaps idea i could have hinges and servos attached to rods pushing and pulling different hinges to change elevation(flap on the back) and left and right(one on each side or on straight up on back like the rudder of a plane), and i do have some experience in flying rc planes, and have succesfully used several flight simulators
-
Yea, basic control surfaces sounds right. Thinking back at the hang gliding thing, you could probably get away with just swinging weight left and right (to steer). You could always just kind of place the weight a bit forward so it's in a constant dive. If you have it in a constant shallow dive you basically eliminate the risk of stalling due to wind, etc.
not to mention it would greatly simplify the robot.
-
yeah, eventually i would like to be able to make something advanced
-
a good idea when you are first starting to experement with this sort of thing is start simple and over time without even realizing it your skills will build nd there for the complexity of what you are able to achieve. ;D
-
I think something like a trike wouldnt be that hard to build... I'd like to see something that can get its self off the ground.... also you could have two motor with propelers on them one on each wing to control tilt(just an idea) solenoids will be too havy and have a pretty hefty price... my best advice is to look into how smaller planes are controlled...(trikes, bushplanes etc.) you may also actually consider a flight simulator
-
Flight simulators will teach nothing at a small scale like this, you will need a RC flight simulator. Mango your idea with servos and linkages is perfect, that is how it is done with rc planes. You are basically building a UAV so maybe check out www.diydrones.com, they are developing the ArduPilot which is very nice and cheap.
-
if you can, go to a model shop and ask to see if you can have a look at a rc airplane ad get them to explain to you the finer details on how the plains work and see if they can recommend any materials and or techniques.
-
even just looking online at RC airplanes (especially electric trainers) and watching videos, sometimes they have demonstrations on how they work (what controls and control surfaces they have)
-
yeah about@com has a ok tutorial on howeach control surface changes the way the plane moves
-
I'm sticking firm by my statement that you should learn to fly RC planes before you start this project. That way you will have a better idea, or even an intuition on what you will need in order for this to even be feasible.
-
i do have experience, i can run a flight simulator(where you are the giant box with hydraulics that move you around, the way people get certified to fly jumbo jets)
-
Flight simulators are absolutely nothing like planes on a small scale like these. The bigger you get, the more stable things are. You really need some in-depth flying with twitchy rc aircraft, because they can go from flying perfectly to on-the-ground in literally the blink of an eye. I will guarantee you will crash/lose your plane without lots of experience! This is particularly why you don't see little UAVs flying around all the time, they require a lot of time and effort to fly correctly. People spend months simply trying to get an autonomous system working on a proven RC airframe, yet alone a new design.
Not to be a downer, but I have seen quite a few people jump right in to UAVs with little RC experience, end up with a hunk of foam a week later and ask why the $400 UAV just crashed.
-
the hydralics is how people learn to fly jumbos but only 10% the rest of the time is with books and in light aircraft. Remeber that you need you pilots licence before you can become a jumbo pilot so they already know how to fly. They just use the simulator to get used to all of the systems as there are thousands and only use it a little bit to work out the handling of airplanes, after that they need to fly the real thing.
-
how is a simulator going to help much? unless its an rc simulator of coarse...
jumbos weigh 1000x what your tiny aircraft will... i think yours may experience a little more turbulence ;D
-
aerodynamics don't always scale. there are lots of RC aircraft designs that wouldn't be workable for a larger aircraft and vica versa. Having experience flying real aircraft doesn't mean much when you try to scale it down. My dad for example was a pilot for years (small single engine planes) and he can't fly an RC aircraft.
-
your wrong... you need a different certification to fly jumbos.... the average pilot spends 2000 hours flying ultralight aircraft before making a transformation to jets. Jumbos are a lot harder to fly becase the bigger they get the longer the delayed reaction becomes though they do have more stable flight... I've heard a hard thing can be getting off the groundin a glider and staying behind the tow and not breaking the tow cable
-
of course i know you need a different qualification to fly a jumbo. But the majority of the learning happens in a class room not in a simulator
-
while most of the learning may appear to happen in a classroom nothing beats experience.
-
Hi,
I'm 98% certain I saw a remote controlled hang glider before, you could look at how they did it (or even buy it and modify it).
Perhaps this: http://www.rokemodell.de/x-int/gb/products/gr/gr4433gb.htm (http://www.rokemodell.de/x-int/gb/products/gr/gr4433gb.htm)?
I vaquely recall another one (from Graupner as well), but it was years, if not decades back in time.
MangoBot <- driving an R/C-car is definitely extremely different from driving a real car (I don't think anyone will disagree on that), so a Jumbo Jet pilot might not even be able to get an R/C-plane off the ground (or land it) without some training.
-
http://www.blac.net/
Hanglider with standard rudder/elevator controls
-
either one of those should be easy enough to modify. no doubt they use servos for control which means it would be very simple to interface a controller to. you could start w/ just having the controller try to fly it straight and have some sort of kill switch to "force" a landing (make it fall out of the sky). or even, at first leave throttle control up to you (leave it connected via rc) and have the robot steer. that seems like it would be a good way to go about it
-
Hate to admit it but I've crashed an R/C plane or two in my time :-\
Flying one is without a doubt not easy to master. The controls can be finicky. Not to mention the factors that are out of your control entirely: The smallest gust of wind can ruin your day with these light models, especially if there is a body of water nearby... *ahem* Not saying I did that ;)
Seriously, you can pick up some really cheap ones at Radioshack to play around with and you won't feel it in your wallet when you crash it the first time. I've never heard of an enthusiast who doesn't have at least one nasty crash story.
Doing so will give you a greater feel for what you're up against, especially when trying to consider scenarios like the gust of wind I spoke of earlier. Think of it like this: It's like taking your drivers test for your license- much easier to do if you've actually driven before and not just read about it or played Grand Turismo a few times.