Author Topic: Flying Robot advice (Read 3183 times)

dual · « **on:** August 17, 2009, 11:21:12 AM »

Hi guys, so I played with a couple robot ideas but were all wheeled. Now I would like to have your opinion on this certain Robot that I want to build. This is the robot that I want to try and immitate. It's called a Haro and it does not actually fly but it glides through air, it's also round so if possible, it should not have rotating blades sticking out.

http://www.youtube.com/watch?v=QKeYl_p-Jug

Basically for now, I just want to try and see if I can let something fly, even if i have it wired to something on the floor (Power for instance) or leave as remote controlled.

These are the ideas I've researched:
1.) Use a helium pod/balloon. If necessary, have fans on the side that will control direction and motion.
2.) Use accelerometers/Gyros to control flight
3.) I know the basic physics is that it has to be super light so I can be able to use a helium balloon to make it float. The idea is very close to a BLIMP

Thanks guys

Finnik · « **Reply #1 on:** August 17, 2009, 11:44:33 AM »

I think you'd have to miniaturize everything down to an incredibly small scale, or make a pretty big balloon, if all you have for lift is the balloon.
I know you'd rather not have rotating blades sticking out, but I think you'll at least need something for direction control like you mentioned.

And I'd say you'd probably come short of lift when component weight adds up, so you'll probably need some rotors for that as well; you might want to have a look at those small toy rc helicopters for inspiration. The other solution to weight problems would be to supersize the balloon, but that might at some point turn to big to be a desirable option.

Altogether, your plan looks alright. I predict that you'll be having numerous weight issues, but go ahead and prove me wrong. I'd love to see this project come together.

dual · « **Reply #2 on:** August 17, 2009, 11:55:40 AM »

Adding a Rotor would not be a problem at all, albeit the extra rotating blades for direction.

However, if those would be necessary, is it possible to enclose the whole system (including the rotor and rotating blades) all together in a spherical body such as what you have seen in the video? Would that constrict the motion? I guess I would have to punch some holes on the spherical body to let the air out or in.

Finnik · « **Reply #3 on:** August 17, 2009, 12:04:31 PM »

Yes, I think you could include them in the orb, but you'll have to ensure proper air flow. I think it would be easiest to fit tubes in the sphere in which you'd mount the propellers, to make sure that it has a proper intake and outlet.

It might also be possible to not have the intake in line with the propeller and outlet, i.e. the intake tube could be bent to fit your needs. Considering airflow, I'd say the outlet better be a straight tube.

guncha · « **Reply #4 on:** August 17, 2009, 12:10:42 PM »

If you make it anything like the thing in the video (which I hope you do), please don't make it as annoying

dual · « **Reply #5 on:** August 17, 2009, 12:12:53 PM »

Seriously.. But yeah, for now, I only want to try and make it "Float" as in the video which is why I am hoping to use slow movements with a helium balloon and try to enclose the whole system in some spherical body.

dual · « **Reply #6 on:** August 17, 2009, 08:52:39 PM »

Say, if I were to enclose my whole system inside an Orb structure, can I eliminate the need for a helium balloon and just use a rotor to level my flight.

I'm worried that the robot might fly forward too fast though, which is why I opted for helium in the first place.

parallax · « **Reply #7 on:** August 17, 2009, 09:36:19 PM »

You may have already thought about this but just in case you hadn't...

Seems to me with a basic shape like a sphere you will need to take into consideration the spinning effect your lift rotor(s) may have on the body of the robot due to torque. The tail rotor on a helicopter acts to counter this... Maybe you could run multiple rotors "against" each other to solve this, as a sphere no inherent stabilizing component in its shape (like a fixed wing aircraft does) nor a tail with a convenient rotor on it (like a heliopter).

Unfortunatley, more motors + peripherals necessary to run them=more weight....
So perhaps some lighter than air gas to offset *some* of the weight? Give it a slightly "negative buoyancy" so that much less power is needed to lift it. It need not be a blimp per say...

If you were to run three or more rotors on the bottom in a tri-pod like configuration all at a 45 degree angle away from the center point of the sphere it might counter the torque effect I was talking about earlier (another idea would be to have the blades set up to run in opposite directions). For movement you could operate all of them at the same time/power for ascending and descending. For lateral motion maybe, it seems to me like if you lessened the power on one "side" of the machine you may be able to achieve motion in that direction through a kind of controlled slide.

What do you guys think?

Like I said, just a thought I had when I was reading your concept which happens to be great by the way.

dual · « **Reply #8 on:** August 18, 2009, 11:21:43 AM »

You're right, I have already thought about that too but in the back of my mind, since the lift rotor I will be implementing should not have to be heavy duty if I also implement a helium balloon then the torque wouldn't have as much effect as the total weight of the system anyway.

But if it does according to your theory, I don't particularly understand the concept of having three rotors on the bottom to counter the torque of the lift rotor. Why does it have to be three? And regarding the positioning, you mentioned 45 degrees away from the center point of the sphere but that would make the spaces between the three rotors uneven. Do you mean to say that they have to have even spaces in between?

parallax · « **Reply #9 on:** August 18, 2009, 06:57:27 PM »

I apologise: I was rather vague in my description

Let me try again but let me preface that with the thought that I might be overcomplicating this

When I said three motors at 45 degree angles, I was referring to the Z axis. If we placed zero degrees at the bottom of the sphere (at the "south pole") 45 degrees would be halfway between zero and ninety (which would be the "equator"). Therefore they would all be working against eachother, but in an upward fashion. As far as their spacing around the sphere All three would be equidistant around the sphere in regard to X and Y: 120 degrees (4pi/3 radians) apart. So to continue my earth analogy, there would be motors aligned away from the earth pushing towards the center from below the equator, spaced roughly at roughly Australia, Africa, and South America.

Does that clarify what I was trying to say? Hope I didn't make a fool out of myself

My reasoning behind three motors is that it would be the minimum necessary to control the sliding motion I described earlier. Lower the power on one of the three, and it will begin to slide in that direction. Also, as each motor is working against the other two per say, it would have the added benefit of eliminating the torque effect that would be produced by one rotor working alone.

Now back to what I was saying about overcomplicating the design..

Quote from: dual on August 18, 2009, 11:21:43 AM

since the lift rotor I will be implementing should not have to be heavy duty if I also implement a helium balloon then the torque wouldn't have as much effect as the total weight of the system anyway.

I believe you may be absolutely right! If you are close to a "neutral buoyancy" anyway you really wouldn't need a strong motor to lift your bot. Hence the torque issue would be a mute point.

What I was describing breaks the K.I.S.S. rule all over the place. What you are describing is a far simpler solution and in fact, is probably much more desirable.

If you wanted to be sure about the torque issue before you started the build, you may be able to calculate if it will be a problem for you if you could see how much force would be necessary to rotate the bot in midair. I would think it would be something like this (although it's been awhile since I sat through a physics class so I might be a little off base

):

In "psuedomath"

F₁ =Air Friction+Inertia
F₂ =Torque generated by rotor

If
F₁ > F₂
You should be all set. The force generated by the torquing effect wouldn't be sufficient to rotate the bot.

However...

If
F₁ <= F₂
You might be in trouble...

What are your thoughts?

Razor Concepts · « **Reply #10 on:** August 18, 2009, 07:18:57 PM »

Three will be tough.. hard to cancel out the torques if the motors are stationary. Either two rotate CW and one CCW, and the craft tends to rotate CW. If two rotate CCW and one CW, the craft will rotate CCW.

Your idea but with 4 motors may work, like a quadrocopter. The tricopters that are currently out there have one or more pivotable motors that stabilizes the yaw.

dual · « **Reply #11 on:** August 18, 2009, 08:05:17 PM »

Yeah, I am not really against the idea of your rotors but the fact that Four rotors over Three motors actually sounds more fault tolerant.

However, there is something I'm not getting though. Granted that I am Computer Hardware Engineer so I'm not as good with Dynamics let alone AeroDynamics,

but... If I do implement a 3 or 4 rotors on 45 degrees from the XY plane and a lift rotor on the top.
Would I still require two other rotating blades for direction?

I think my system will be too heavy in this case.

Razor Concepts · « **Reply #12 on:** August 18, 2009, 08:17:13 PM »

Nope, 4 rotors enough for all directions.

parallax · « **Reply #13 on:** August 18, 2009, 08:24:15 PM »

That makes a lot of sense, and simplifies the physics of it. Odd to think that by adding more motors you can actually simplify the problem!

I just found this a moment ago:
http://en.wikipedia.org/wiki/Quadrotor
Seems to have a lot of pertinent information about the basics of four rotor control.

dual · « **Reply #14 on:** August 18, 2009, 08:38:42 PM »

Interesting system. However regarding direction - Forward, Left and Right - How complicated would it be to control the speed and direction for four Rotors arranged as such compared to a my original system plan that has one lift rotor, two direction and a helium balloon?

If it wouldn't be as complicated then I might just have to implement the four rotors.

However I've also seen a design that is actually also in the form of a sphere.
It has two big rotors, one on top rotating against the other on the bottom that cancels each other's torque effect. Would that actually work?

parallax · « **Reply #15 on:** August 18, 2009, 09:16:12 PM »

It seems to me that with your original concept, as long as you increased the thrust of one of your directional motors in the proper proportion with an increase in thrust from your lift rotor you would be okay. For instance, when you would want to increase in altitude, your lift rotor would generate more torque with the increase in power. For the sake of argument let's say its in a clockwise direction. If you increased the thrust of your directional motors in a counter clockwise direction in a directly proportional amount to the increase in thrust from the lift rotor the forces would cancel each other out.

From what I gather about a quadrotor design (and I'm by no means an expert), control can be achieved by simply decreasing/increasing the thrust of pairs of rotors mated by their axis.

To summarize:
Your concept requires would require knowing exactly how much rotational velocity would be generated by an increase in power to the lift rotor, and countering that exactly with your directional rotors... A math and programming issue.

A quadrotor concept would require coordinating pairs of thrusters to act together. Less math and inherent torque issues perhaps, but a more complicated hardware setup.

As far as the two rotor concept you are describing- has it been done? If so can you post a link? I'd love to read more about it

First thoughts are that in order to increase/decrease in altitude one rotor would have to be operating at a higher thrust than the other, which would generate torque. Also, how would you achieve movement in a horizontal fashion? Tilting the rotors perhaps?

dual · « **Reply #16 on:** August 18, 2009, 09:21:56 PM »

This design is generally what I'm hoping to carve mine into:

http://www.surveyor.com/YARB.html

The main difference is the helium balloon here is big enough that it would not require the help of a lift rotor.
The two direction rotors are further from the center of mass and stretched by a shaft I pressume.

As far as I've understood, the concept behind the YARB is the helium balloon has enough bouyancy to maintain the lift and the directional rotors mainly act like two wheels on a wheeled robot, rotating clockwise or counterclockwise based on the direction - forward, left, right.

parallax · « **Reply #17 on:** August 18, 2009, 10:05:30 PM »

Thank you Dual- that clarified the concept for me greatly. I love what the creator has done with the tilt-control sensors in the phone!

A lift rotor would definitely facilitate the sphere shape you're going for and keep the size of the bot smaller in regard to the helium balloon. As far as sticking to the Haro idea and keeping the rotors somewhat enclosed rather than sticking out, you may have luck hiding them within the gondola as long as your intake/outtake channels are large enough to not obstruct airflow to badly. As long as the gondola's outer shell is of a light enough material, you should be able to make it a large enough size to do this and not have it be to obvious. The lift rotor might be a bit more of a challenge to hide...maybe you could form fit the balloon around its intake/outtake tube?

For what you are trying to accomplish and given your inspirations for the robot (Haro and YARB) I think your concept works beautifully, and a quadrotor design would be overcomplicating things

dual · « **Reply #18 on:** August 18, 2009, 11:35:31 PM »

Yeah, I think for now I have most of the concepts/ideas that I need to get started. What I have to face now is researching the material for the Airflow tubes that are bendable at my will, Spherical shell (will probably be just some toy from a toy store) and the hardest thing, a helium balloon that can provide me with the necessary bouyancy.

And since this is not supposed to be a professionally manufactured product anyway, it does not have to be so small as to look exactly like the one in my video and the rotating blades are not really restricted from being within the spherical structure.

I, like all of us here, just want to have my fun building my first dream robot!

dual · « **Reply #19 on:** August 25, 2009, 11:01:27 AM »

So, I think using a helium balloon would not work after all. I was looking at a 36 inch, 4 ft^3 helium contained saucer type blimp from RC toys, and the density of helium and air.

After calculations, I found that at ground level and 25 deg celcius, the 36 inch helium balloon can only carry .25 pounds. I maybe wrong on this but it still makes sense after thinking thoroughly.

If I still implement a lift rotor, how fast should I run it? If the weight that I want to carry say is 1.5 lbs? If someone is willing to answer?

Razor Concepts · « **Reply #20 on:** August 25, 2009, 11:04:27 AM »

1.5lbs is quite a payload for something like this. But you will only have one lift rotor? There will be huge torque cancelling issues witht hat.

dual · « **Reply #21 on:** August 25, 2009, 11:06:40 AM »

1.5 is actually only an estimation. I will have a PCB board, 1 lift motor, 2 forward rotors cancelling the torque, 1 camera, 3 sonars, accelerometer.

parallax · « **Reply #22 on:** August 26, 2009, 09:26:07 AM »

Did you have specific components in mind? Their spec sheets should give you the exact weights.

Something doesn't quite add up- The YARB has a very similiar setup to what you are trying to accomplish, and yet it doesn't have a lift rotor at all. Correct me if I'm wrong (I'm horrible at judging size unless I'm right next to something) but it has a similar size balloon to what you are describing. Yet, it seems to have a positive lift to weight ratio (notice in the video that in order for it to descend the rotors need to tilt downward).

Have you double checked your math on the lift? How close of an estimate is 1.5lbs? Infield? Ballpark? Parking lot?

Seems to me the answer is in there somewhere.

News:

Author Topic: Flying Robot advice (Read 3183 times)

Related Topics