pic: All Terrain Swerve 1

[Alex.Norton]

loading photo ...

[GMAdan]

I have to say i am impressed nice design both in the cad department and the actual design it is great. Only one question, do you think having the cim motors mounted so low will become a problem? if you do go on rough terrain I believe they will come in contact with the terrain a lot.

[AChastain]

It is indeed impressive. As far as the CIM motors go, I can see what GMAdan is saying. Any fall could knock 'em loose then you could be literally stuck on a mountain.
The other thing is that if you ever hit any flat patches, it seems like those treads would have trouble turning. Four treads pulling a "bunny hop" comes to mind. Ahh, that could be interesting.
Other than that, it looks like a sound idea. I'm no expert but I say go for it.

[1359th Scalawag]

It looks really cool!

I've never thought about doing 4 little tank drive systems before. Was this a design you used for an actual robot you used in competition or is it just an idea?

The only thing I could see that looks like a problem would be the placement of the cims. Might the cims on the left be better on the inside too? When driving by something, they look like they might be hit.... I dont know. I have no experience in that area.

Wow. Very impressive. One suggestion: You're going to have trouble turning. I suggest using the Mini Bike motor through a really torqued out gearbox to steer, as those won't be able to turn that thing.

Very nice!

[bjimster1]

Looks to me as if each module rotation point are going to be very hard to keep from breaking. I think theres too much leverage under the rotation point, maybe im wrong.

way to think different though.

[=Martin=Taylor=]

I think your real turning problem is going to be draaaaging those high-traction treads around on a smooth surface.

Why not reduce the area of contact between tread and floor?


This was an old idea from 2004 (see thread >here). The treads had only a small area of contact with the floor - but a large area of contact with oncoming obstacles.

[Alex.Norton]

Thanks for all the interest

I admit that the CIM's are very close to the ground, however, the treads aren't very wide at the moment and it would be easy to go or 3 inch treads instead of one and a half inches. As for the CIM's being on the outside that was just a mistake on my part and I will fix that soon

I have thought a lot about the turning problem and had a couple of ideas. I thought about using another set of motors on each module to lift it up onto just one of the two pulleys. I also thought about making the shock absorbers to out of pneumatic cylinders, and put valves on it so that it could lift the modules onto one pulley. The other advantage of these is that it would make it easier to get up ledges.

I do like the idea of using three pulleys to get one point of contact but I'm weary of this because it wouldn't be stable in this set up with the second axis of turning. This is actually just a fun little CAD project that I'm working on. I'm trying to put together a bunch of different wierd drive trains together to give the younger kids on the team some ideas for next year.


Again thanks for all the input

[Stephen Kowski]

Originally Posted by Hachiban VIII

This was an old idea from 2004 (see thread >here). The treads had only a small area of contact with the floor - but a large area of contact with oncoming obstacles.

I don't think this was an old idea from 2004, but rather an idea from an old robot in 2001 (http://www.chiefdelphi.com/media/photos/12374).....I think some of you might know the team....

[artdutra04]

So far this looks like a really neat idea.

The only problems I see that might be a problem with going offroad and all-terrain are the lack of ground clearance between the CIM motors and the ground, the "exposed" gears on the outside of the modules (although protected from the side, there's nothing on the bottom to prevent sand or grime from getting in there), and maybe the treads themselves.

I'd probably consider much wider treads with more traction to them, so that if traveling through something like sand, snow, or mud they don't sink out of sight. (Ever try to ride a bicycle through deep mud?) Wider treads will spread out the weight of the robot better, allowing for the robot to glide over the obstacles instead of sinking into them.

Originally Posted by Alex.Norton

I have thought a lot about the turning problem and had a couple of ideas. I thought about using another set of motors on each module to lift it up onto just one of the two pulleys. I also thought about making the shock absorbers to out of pneumatic cylinders, and put valves on it so that it could lift the modules onto one pulley. The other advantage of these is that it would make it easier to get up ledges.

I do like the idea of using three pulleys to get one point of contact but I'm weary of this because it wouldn't be stable in this set up with the second axis of turning. This is actually just a fun little CAD project that I'm working on. I'm trying to put together a bunch of different wierd drive trains together to give the younger kids on the team some ideas for next year.

Point contact would a complete 180 of what you need when going off-road and all terrain. You need more contact with the ground, to better spread out the load so you don't sink into the ground.

Originally Posted by 114ManualLabor

One suggestion: You're going to have trouble turning.

No it won't. This robot should have an excellent time turning, and I don't see any problems with it as is. In fact, with some sophisticated code it can be even more agile than a traditional swerve drive in it's ability to do complex curves and arcs without any side scrub.

All of the four swerve modules are all independently driven, which means you can have all your wheels turn in different amounts to almost perfectly turn using Ackerman steering. Since every wheel can turn independently, each wheel can turn the exact amount needed to eliminate any and all side scrub. Here's a quick illustration:



...Which would work exactly the same as the Jeep Hurricane concept car:


http://video.google.com/videoplay?do...03950591957346

[bear24rw]

http://www.robots.org/images/NASA_Am...yMarsRover.jpg

[SamC]

Originally Posted by artdutra04

...Which would work exactly the same as the Jeep Hurricane concept car:

...I want one!

Originally Posted by artdutra04

No it won't. This robot should have an excellent time turning, and I don't see any problems with it as is. In fact, with some sophisticated code it can be even more agile than a traditional swerve drive in it's ability to do complex curves and arcs without any side scrub.

Actually, if the motor couldn't even spin the tread (which it wouldn't, those things have practically no tourqe) then being able to do those complex curves wouldn't really matter. I was referring to the motor choice to spin the actual module, not the over robot turning ability.

[Alex.Norton]

I don't see how those motors don't have much torque. The CIM's have more than enough power for the drive on a 40-60 pound bot and the fisher price through a 256 reduction should have like 60 ftlbs. of torque which should be more than enough.

In any case the robot should be moving when turning the modules which will significantly reduce the turning force

[David Guzman]

Originally Posted by 114ManualLabor

(which it wouldn't, those things have practically no tourqe)

I think you need to work out your numbers again. The motors have quite a bit of torque. We used them to pivot our arm, which takes more torque than turning those modules. The only problem I do see with this motors is overheating, but thats a different story. In the end we changed them for globes but only because of the overheating problem. They had sufficient of torque.