Omni Drive Idea

[Jared]

After reading through the discussion in Team 67's CAD thread about getting out of t-bones, I had an idea for a t-bone escaping omni drive that's simpler that has good traction, two speeds, and only requires two single speed gearboxes and was wondering if it was feasible/useful. It does have a drawback- it can't turn in high speed (but it goes sideways).

It is very similar to a normal 6WD drive, except there is an extra set of back omni wheels that have 1/2" ground clearance and the normal traction wheel is moved forward to make room. When sitting normally, it drives exactly like a 6WD. This set of omni wheels are driven exclusively from the left gearbox.

On the front of the robot, there are wheels sticking out the front going sideways. They are in a normal 1"x2"x.125" wall box, but this box can rotate so that the wheels can rotate up and out of contact with the ground. They are driven through bevel gears (or worm?) through the rotating pivot. They are driven exclusively from the right gearbox. When the wheels are rotated down into contact with the ground so that they are are perpendicular to the ground, the front of the robot is raised up off the ground an inch or so. Now, the robot is on the front sideways facing wheels (omni wheels) and the back pair of omni wheels. When the left gearbox is driven, the robot moves forward/back, and when the right gearbox is driven, the robot will move left/right. You could possibly set these gear ratios to be faster than the 6WD so the omni wheel configuration will act like a high gear.

Like this sketch, but with the big omni on the inside of the frame.

http://i.imgur.com/HGrSeNq.png

[theawesome1730]

We did something really similar but not quite like you have described. We had a pneumatic cylinder that pushed down a ball bearing roller into the carpet and we have powered omnis on the front of the 6 wheel drop. Our driver seemed to like it and it was pretty effective at getting us out of friction locks.

Go to CD Media and search for the term "Butterfly Drive". I think you'll find some useful information for your endeavors. If not message me and I can get you it.

[Jared]

Originally Posted by Andrew Lawrence

Go to CD Media and search for the term "Butterfly Drive". I think you'll find some useful information for your endeavors. If not message me and I can get you it.

I've seen butterfly drives before. I've done some more research, and this is more like 1625 in 2011 with the rotating front piece (http://www.chiefdelphi.com/forums/sh...d.php?t=92650), but with omni wheels. The advantage is when you put down the front wheels (which will be omni wheels), the back omni wheel touches the ground, and you can go forward/back and left/right; you don't have to pick. The front wheels are driven by the left gearbox, and control strafing, and the back omni wheels are driven by the right gearbox, and control normal forward/back. You can't turn in this mode. You don't need a third gearbox to get strafing motion though.

Originally Posted by Jared

I've seen butterfly drives before. I've done some more research, and this is more like 1625 in 2011 with the rotating front piece (http://www.chiefdelphi.com/forums/sh...d.php?t=92650), but with omni wheels. The advantage is when you put down the front wheels (which will be omni wheels), the back omni wheel touches the ground, and you can go forward/back and left/right; you don't have to pick. The front wheels are driven by the left gearbox, and control strafing, and the back omni wheels are driven by the right gearbox, and control normal forward/back. You can't turn in this mode. You don't need a third gearbox to get strafing motion though.

I see what you mean, though how is this better than 1625's Lobster drive then? And furthermore, why do you need to go sideways at all? Does it serve a specific purpose in a general strategic analysis you planned when designing this drive?

[asid61]

Interesting concept. It's kind of a like a butterfly drive, but the back omni wheels are sideways. I think if you design a drivetrain with this in mind it could work fine geometry-wise. You would need to do some creative work in order to keep the size acceptable.

I have a quick question: if only the back omnis are sideways, how cna you go sideways? You wil just turn in place, as the back of the robot will swing sideways while the front stays still or moves forward. Drifting would be super cool with this, but I'm not sure about strafing.

[nicholsjj]

When I read this thread today another thread quickly came back to mind: http://www.chiefdelphi.com/forums/sh...967+drivetrain
This should help give you all some inspiration on a drivetrain concept almost if not identical to yours, except it sound as though you want to mix things up a bit. Someone from 967 is WAY more qualified than myself to tell you how well this drive train can be. Also I am a big fan of what 1730 did with their ball caster/omni-wheel set up and I know an 8" wheel version of that is being talked about as a prototype drivetrain for our team next year.

[Jared]

Originally Posted by Andrew Lawrence

I see what you mean, though how is this better than 1625's Lobster drive then? And furthermore, why do you need to go sideways at all? Does it serve a specific purpose in a general strategic analysis you planned when designing this drive?

So far, it's just an idea I had when I was reading the discussion about how team 67 would get stuck in pins. I know lots of teams (118, 1114, 1730) had drop down omni wheels that would let them get out quickly, and I thought it would be a neat idea to mount the drop down omni wheels sideways, and make the driven, so you could drive out of a pin easily. Also, if you're trying to line up to do something, and you're a little bit misaligned, you can easily slide sideways instead of backing up and having to realign.

It's better because when it has articulated down, you get omni directional movement, and possibility for a higher gear ratio when traveling in both directions. It also only has one side with articulating wheels, so it's simpler to build.

With only two single speed gearboxes and one pneumatic actuator, you get two speeds, the ability to turn, or have crab drive (translate in any direction). You pick from low speed normal 6WD driving and high speed crab drive.

Originally Posted by asid61

Interesting concept. It's kind of a like a butterfly drive, but the back omni wheels are sideways. I think if you design a drivetrain with this in mind it could work fine geometry-wise. You would need to do some creative work in order to keep the size acceptable.

I have a quick question: if only the back omnis are sideways, how cna you go sideways? You wil just turn in place, as the back of the robot will swing sideways while the front stays still or moves forward. Drifting would be super cool with this, but I'm not sure about strafing.

The front is the side with the sideways facing omni wheels.

The back will be on omni wheels too, so I'm hoping it will slide sideways.

I think if I move the rotating set of omni wheels to the middle, it will work much better. When they rotate down, the robot will tilt, and the back omni wheels (facing forward) will come into contact with the ground.

[Adrian Clark]

Originally Posted by Jared

The back will be on omni wheels too, so I'm hoping it will slide sideways.

Unfortunately I don't think this will be the case.

I suggest you look at the thread on 967s similar drive, there's a link a few posts back. Their setup is pretty much the same in terms of function. When they strafe they do a turn with the omnis on the ground to counteract the turning force from the sideways wheel. For your drive train you would need to do something similar to properly strafe. Unfortunately, if I understand your design correctly, this is not possible with your drive as the back set of omnis is driven by the same gearbox.

-Adrian

[RyanCahoon]

Originally Posted by Jared

It's better because when it has articulated down, you get omni directional movement, and possibility for a higher gear ratio when traveling in both directions.

I think if I move the rotating set of omni wheels to the middle, it will work much better. When they rotate down, the robot will tilt, and the back omni wheels (facing forward) will come into contact with the ground.

I was playing with a similar drive idea a couple years ago (about the same time as the Iron Lions were doing their drive).

Here's an Inventor animation of the shifting action, and here's a VEX prototype of the drive. Even though it's more severe in the VEX prototype than it would probably have to be on a well-machined FRC bot, the biggest problem I saw with these designs is the amount of tilt necessary to get the traction wheels off of the ground, while still maintaining a decent wheel base in traction mode (these are competing interests).

The Iron Lions' design circumvented this by making the omni wheels part of the support in traction mode, but this means you can't get the dual gear ratios out of the switching system.

It might be interesting to look into utilizing 148's strafing wheel design from this year, where the strafing drive motors also do the actuation of the strafing wheels.

[Bryan1625]

This year we had a single speed 6 CIM drive with 2 traction wheels on the back and 2 omni wheels on the front and T boning was never an issue and if we wanted to T bone someone our front end stuck on them because of the omnis. Its simple cheap and very effective