pic: My wheel/leg design

[mechanicalbrain]

loading photo ...

GREAT! Just a thought ...
you'll need some pretty sophiticated stuff to be able to stop the robot.
In the show the robot kinda turn and difts to a stop

[Kevin Sevcik]

From a more practical standpoint, if you make that entire shaft spin, you'll be putting a terribly large load on the bearings and shaft. Ball bearings are intended to support mostly radial loads. With your current design, the shaft bearing at the top would see a large thrust load, which would drastically shorten the bearing's life if you don't over-spec the bearing by a lot. Also, keep in mind that flexible shafts aren't going to transmit much torque.

[greencactus3]

i think a simple doublewishbone idea of a suspension would work just as well and simpler.

[mechanicalbrain]

Originally Posted by greencactus3

i think a simple doublewishbone idea of a suspension would work just as well and simpler.

I specifically want a simple walking/rolling design.

Just wondering but, could you explain how it works? It look pretty cool

[mechanicalbrain]

Well the general idea is that their are four legs and the sit out at an angle from the robot. The entire leg can rotate on that gear (more symbolic of where it pivots then anything else). Also the rod is attached to that square coupler by a ball bearing and rotates. When it rotates the half sphere at the bottom acts like a wheel. Its just a conceptual design right now and not really possible due to the limits of last years kit but its a future plan. Not only could it have simple walking but the legs could start facing up and fold down to add significantly to the height of a robot. Basically I'm trying to come up with a simple walking robot that doesn't have the disadvantages of speed that most walking robots posses.

As mentioned its got a few engineering hurdles it needs to overcome but if i don't use something like this in FIRST then ill definitely do it privately because the idea has been bouncing around my head for quite some time i just never really developed it until now.

[phrontist]

In what way does this "walk"? Iit just a wheel at the end of a shaft? Or can it also swing the shaft some how...

[Kevin Sevcik]

It "walks" by whipping around that large gear the shaft bracket is attached to. Mind you, it'd walk better if the leg moved in and out as well as around or something similar. Perhaps instead of starting from scratch on both counts, you should find some simple leg joints that work for people and then figure out how to put wheels on them?

[mechanicalbrain]

Originally Posted by Kevin Sevcik

It "walks" by whipping around that large gear the shaft bracket is attached to. Mind you, it'd walk better if the leg moved in and out as well as around or something similar. Perhaps instead of starting from scratch on both counts, you should find some simple leg joints that work for people and then figure out how to put wheels on them?

Yes thats exactly how it works. As to the retracting part. It would have to involve a mechanical function so i can keep the motors to a minimum. I want it to come around back so it can clear stairs and such objects. Finally, If I added wheels to the end of legs it would need to be some type of pulley system and then again I run into having more then two motors a leg.

[artdutra04]

I you want to see a ton of different walking robot designs, check out this page. (You'll have to scroll down a bit.) I know that these are Lego creations, but the principles behind the walking robots can be adapted to a larger scale. IMHO, I would think that instead of rotating the entire leg around, do as Kevin Sevcik suggested, and make the legs go up and down, instead of in circles. Similar to this picture:


I know that it wouldn't be able to go up stairs, but it would have much less stress on the shafts.

Also, I am curious. What is the size of your leg?

[mechanicalbrain]

Originally Posted by artdutra04

IMHO, I would think that instead of rotating the entire leg around, do as Kevin Sevcik suggested, and make the legs go up and down, instead of in circles. Similar to this picture:

Thanks for the link! Ill go through it. You see the point is that Im TRYING to do a wheel/leg and not just a walking design. Your typical walking robot is very, very slow. I want to have some type of wheels to give it a good speed and then legs to give it the best climbing available while using a minimum amount of motors. In answer to that last question, I was looking at making the legs 1' - 1.5'. Thanks agian for the link.

[Matt Krass]

You may want to try a tri-wheel.

Essentially its three wheels centered around a point, 120 degrees apart at the center of each wheel. Two rest on the ground and all three spin. When you hit a stair well, the friction cause it to pivot around the center and climb up the step, in theory. I played with the idea but never built it, I'm gonna see if I can do it with the Vex kit I'm getting this weekend.

EDIT: I added a rough sketch.

[artdutra04]

Originally Posted by Matt Krass

You may want to try a tri-wheel.

I've built one of those before using Legos, and it worked pretty well. But there is one thing to remember when you build these. The wheels around the outside must spin faster than the center shaft. This will make the three wheels roll easily, but when an obstacle comes (like a stair), the force will then take the easier route and rotate the whole assembly (instead of the indiviaual wheels).

If you are interested, here is video of a Lego Tri-Wheel robot climbing stairs: http://www.visi.com/~dc/tristar/stairs4.avi

There are more pictures on this page:
http://www.visi.com/~dc/tristar/

[Travis Hoffman]

Got whegs?

http://www.chiefdelphi.com/forums/sh...ighlight=whegs

Courtesy of researchers at Case Western Reserve University in Cleveland, OH.

Have fun putting wheels on the ends of these things!