Square Wheels?

[mechanicalbrain]

Well this is definitely something I would have never considered but in retrospect makes allot of sense. square wheels actually give an overall circular rotation to the robot. My question is this, anyone interested in making a square wheeled FRC robot? Also I wonder how well this would work on stairs?

[sanddrag]

That would be one bumpy drive. I suppose it might have some sort of advantage on stairs (only guessing here) but not a lot, and especially not made out of something like the slippery plastic shown in the picture. It would take more torque to get moving than with round wheels.

For climbing stairs, I think whegs are a good way to go. Except they require lots of torque, and the performance on flat surfaces is sacrificed.

[KenWittlief]

Triangular wheels would be a 25% improvement over square wheels.

For each rotation, there would be one less BUMP !

[sanddrag]

Originally Posted by KenWittlief

Triangular wheels would be a 25% improvement over square wheels.

For each rotation, there would be one less BUMP !

And linear "wheels" would be a 33% improvement over triangular wheels. Only two bumps per revolution!

[KenWittlief]

So mathematically, the best wheel design is a single arm: one bump per revolution.

BTW, goto ebay and search on "matt mason space crawler"

a very popular toy back in the late '60s

[i_am_Doug]

Originally Posted by sanddrag

And linear "wheels" would be a 33% improvement over triangular wheels. Only two bumps per revolution!

Oval wheels....One bump per Revo.

[Elgin Clock]

I think the way square wheels would be useful in theory anyways (my theory that is.. lol) is the fixing of the instance you get when you have your motors lag behind each other. Not taking sideways vibration into account, the only way you can move with square wheels is 4 times or 4 bump movements. I think if it is regulated enough with traction, that you can move forward equally between the motors in reverse directions on the sides of the bot.

Obviously, taking sideways vibration into account, the motion won't be absolutely straight, so maybe this post is just a useless one, but think about it at least.

[Eric Scheuing]

Square wheels seem like a good idea at first, but with this years game, you'll probably need the camera, and if there is any sort of movement like that at all, the camera is going to have some serious problems tracking the goal, not to mention shooting.

[Arkorobotics]

why in the world would you want to do that?

[mechanicalbrain]

Actually from reading the article a I understand it their is very little bump at all. It has to do with the fact that their is a rotation of corners toching the ground. Interesting concept really. It reminds me of how whegs work.

[Madison]

Originally Posted by mechanicalbrain

square wheels actually give an overall circular rotation to the robot.

Square wheels will not provide rotation like that of a disc. The center of rotation does not move linearly, but in a cam motion -- a bit like a spiral, but not repetitive. Whegs, which I assume is a combination of wheels and legs, might appear to create circular motion as well, but in reality would form a similar cam-like path. The whegs simply rotate fast enough to minimize the time the machine has to fall between legs. The faster they spin, the less 'bump' their appears to be.

The 'square-wheeled' machine moves due to a shift in its center of mass and thus, requires specially designed wheels to the shift of mass causes each corner of the robot to fall sequentially.

I can't imagine how this would be useful to a FIRST machine at first glance

[mechanicalbrain]

Originally Posted by M. Krass

Square wheels will not provide rotation like that of a disc. The center of rotation does not move linearly, but in a cam motion -- a bit like a spiral, but not repetitive. Whegs, which I assume is a combination of wheels and legs, might appear to create circular motion as well, but in reality would form a similar cam-like path. The whegs simply rotate fast enough to minimize the time the machine has to fall between legs. The faster they spin, the less 'bump' their appears to be.

The 'square-wheeled' machine moves due to a shift in its center of mass and thus, requires specially designed wheels to the shift of mass causes each corner of the robot to fall sequentially.

I can't imagine how this would be useful to a FIRST machine at first glance

sorry I misread part of the article. The idea behind this is not just the square wheels but the fact that no motor is directly driving them. All motion of the wheels is accomplished though shifting the weight of the robot and a little help from gravity. The idea applies in that you can control multiple objects (and with more things then just wheels) without direct contact. Imagine designing a drive train that changes gear ratio when tipped. Sorry about the fixation on the wheels.