Chief Delphi - 6 Wheel Drives Without Dropped Centers

Quote:

Originally Posted by cmrnpizzo14 (Post 1274720)

I am seeing a lot of conflicting posts about versa wheels so could someone clarify a little bit? As far as I know the most defining characteristics are:

1.) Great traction in the standard forward backward drive directions
2.) Poor traction if robot is being pushed sideways while parallel to floor
3.) Great traction if robot is pushed sideways but tips up so that the treads "dig in" to the carpet

If anyone could confirm or deny these that would be very helpful. And if these are true then it appears that these would be perfect for 6wd w/o a dropped center since they should have minimal scrub due to their poor lateral friction.

Here's some video evidence. At 0:54, 2834 hits 469, accelerating from about 4 feet away, and shoves 469 about 3 feet. 469 uses VersaWheels. (I got a good look at their drivetrain Saturday evening, about two weeks ago.)

As to your question about their use in the corners of a no-drop drive, I'll reaffirm what I said in the other thread, that I don't think that's a great idea. Consider a 6-wheeled no-drop robot with VersaWheels in the corners. For simplicity, assume the drive base is a perfect square with the horizontal center of mass at the center of the horizontal plane. Thus, the turning circle should touch the contact points of the VersaWheels. Now think about how the tread interacts with the turning circle. The tread looks like this:

-----------
~~~~~~
-----------
~~~~~~
-----------

more or less. And the turning circle at the patch, when superimposed on the tread pattern, looks like this:

-----------/
~~~~~/~
------/-----
~~/~~~~
-/----------

sort of. (Imagine the slashes as being a continuous curve.) When the wheels exert force on the ground, the easiest way for them to slide is sideways. But they instead want to slide along the turning circle! This means a significant portion of the force is wasted trying to turn against the "W" tread.

Of course, this is all theory. The best way to test this kind of drivetrain is to build it. I just don't predict that it'll work out too well.