Drive Train Prototype [Archive]

nitbaj

23-06-2010, 22:41

Team 461 has been prototyping an omni-directional drive train concept we call Jump Drive. It switches between mecanum wheels and traction wheels to allow omnidirectional movement and also pushing power (similar to 148's "nonadrive"). We began building it in the Fall of 2009 but were only just able to complete it and get video.

Hope you enjoy it and please let us know what you think.

Video: http://www.youtube.com/watch?v=uEUIshtjbe4

Pictures: http://boilerinvasion.org/index.php?page=photos&g2_itemId=24653

Hawiian Cadder

23-06-2010, 23:56

how much does that weigh?

548swimmer

23-06-2010, 23:58

how much does that weigh?

Done properly, under 40 lbs.

sanddrag

23-06-2010, 23:59

Reminds me of team 980 in 2003, but they did it for switching "gears" not directions.

Brandon Holley

24-06-2010, 00:02

Very cool idea. I like the innovation.

As is, how much does that weigh? (not how much will it weigh eventually...)

-Brando

Hawiian Cadder

24-06-2010, 00:03

something i thought up that could either be more complex or simpler, a tank drive, with only omni wheels and some method to prevent sideways slip, that turned into a holonomic drive by angling the wheels. i think that because there would only be one set of wheels, it might leave more interior room for electronics and what have you.

CENTURION

24-06-2010, 00:04

WOW! Now that is cool :)

I kinda don't like how it bobs up and down when it switches wheels, but it's not really a problem (though it could easily be fixed with a few omnis, or casters)

GarrettF2395

24-06-2010, 00:08

Darn it! You guys beat me to it.
I've been playing around with this idea in my head for awhile now, though it's great to see someone actually build it.
Ok question time.
Is there a reason as to why you decided to have the mechanums on the outside instead of the inside?
Did you guys play around with any other configuration ideas, such as having all 8 of the wheels on the ground at once?

Great work, make sure to run some good tests on it and let us know what you find!

Andrew Remmers

24-06-2010, 01:01

How many Solenoids are there?

Chris is me

24-06-2010, 01:07

Done properly, under 40 lbs.

Where did you find out how much 461's prototype weigh? Or any mecanum - traction combo, anyway?

I heard (if you're wondering where, in the pits at Atlanta) the 148-217 drivetrain this year, while it had an extra wheel, weighed a lot more than that.

Ether

24-06-2010, 07:09

Where did you get those mecanum wheels? Never seen ones like that before.

~

Jared Russell

24-06-2010, 07:35

Where did you get those mecanum wheels? Never seen ones like that before.

~

I believe they are these (http://www.andymark.biz/am-0137.html).

548swimmer

24-06-2010, 09:55

Where did you find out how much 461's prototype weigh? Or any mecanum - traction combo, anyway?

I heard (if you're wondering where, in the pits at Atlanta) the 148-217 drivetrain this year, while it had an extra wheel, weighed a lot more than that.

For Overdrive we had a robot with eight wheels that could switch between two sets of four. The chassis sitting in our shop weighs under 40 lbs, though barely.

Alan Anderson

24-06-2010, 09:56

Does it spin in place well? It looks like the mecanum wheels might need to be swapped.

Ether

24-06-2010, 12:08

I believe they are these (http://www.andymark.biz/am-0137.html).

Ah. Looks like they've been re-designed. The roller mounting tabs look a lot sturdier. It looks like the rollers still rub against the mounting tabs though. That's a bit disappointing.

~

Chris is me

24-06-2010, 12:13

Ah. Looks like they've been re-designed. The roller mounting tabs look a lot sturdier. It looks like the rollers still rub against the mounting tabs though. That's a bit disappointing.

~

I think that's an intended feature of the wheel; you can tighten and loosen the rollers to get varying degrees of forwards versus sideways traction.

Ether

24-06-2010, 12:43

I think that's an intended feature of the wheel; you can tighten and loosen the rollers to get varying degrees of forwards versus sideways traction.

So it's not a bug, it's a feature? :-)

Seriously though, even if you loosen the rollers, they are always end-loaded when the wheel is being driven. This creates substantial friction. Roller bearing friction increases traction in the fwd/rev direction, but decreases traction and force in the strafe direction. For many applications (like Breakaway?) this may be a good trade-off.

I like the look of the new steel end-plates. The mounting tabs look quite a bit more rugged. The aluminum tabs on the 8" AM mecs bend very easily - even to the point of locking up the roller.

~

jamie_1930

24-06-2010, 13:11

Darn it! You guys beat me to it.
I've been playing around with this idea in my head for awhile now, though it's great to see someone actually build it.

Great work, make sure to run some good tests on it and let us know what you find!

That's pretty much what what I was going to say. I had already started CADing this and was trying to think of some other methods to switch it than pneumatic cylinders, but nonetheless I love it, keep up the good work guys.

Alan Anderson

24-06-2010, 14:40

Seriously though, even if you loosen the rollers, they are always end-loaded when the wheel is being driven. This creates substantial friction.

I don't understand. Are you stating that as a deficiency of this specific wheel design, or of mecanum wheels in general?

rutzman

24-06-2010, 15:20

AdamHeard

24-06-2010, 15:24

For Overdrive we had a robot with eight wheels that could switch between two sets of four. The chassis sitting in our shop weighs under 40 lbs, though barely.

I'm not sure what you're referring to, but when my team quotes drivetrain weight we talk EVERYTHING; wheels, electronics board, battery mount, chain, bolts, etc...

If you hit 40lbs with that standard, I'm highly impressed.

electron

24-06-2010, 15:43

Darn it! You guys beat me to it.
I've been playing around with this idea in my head for awhile now, though it's great to see someone actually build it.
Great work, make sure to run some good tests on it and let us know what you find!

I must say that I too have been toying around with this idea in my head. impressive work, though. I hope you guys get to use it in next years game :)

Bjenks548

24-06-2010, 16:24

As far as the 40lb one from the robostangs:

I haven't weighed it myself but heres a picture of it. Imagine changing the larger wheel for a mecanum and mirror those wheels in the front. http://www.chiefdelphi.com/media/photos/27459

548swimmer

25-06-2010, 00:00

I'm not sure what you're referring to, but when my team quotes drivetrain weight we talk EVERYTHING; wheels, electronics board, battery mount, chain, bolts, etc...

If you hit 40lbs with that standard, I'm highly impressed.

It's about 40 lbs with all the wheels, the old IFI control system, battery MOUNT but no battery, and all the pneumatics.

Andy Baker

25-06-2010, 00:18

Way to go Nitesh and WBI - this is a cool design! I hope you bring this to IRI so we can check it out.

I like the look of the new steel end-plates. The mounting tabs look quite a bit more rugged. The aluminum tabs on the 8" AM mecs bend very easily - even to the point of locking up the roller.

~

The bodies of these wheels are actually old and an inferior version of the current version. The plates used on this jump design are from 2008. These tabs were significantly less rugged, since the steel wall thickness was only 0.04" instead of the current thickness of 0.05".

I find that the 8" Mecanum tabs are quite sturdy.

Andy B.

nitbaj

25-06-2010, 09:59

how much does that weigh?

I currently weighs around 90 pounds with everything on it. It's pretty ridiculous, but we didn't really worry about weight when designing it this time around. The frame members are pretty much all 80/20 and we didn't put lightening holes in any of the other plates, so I imagine the weight could drop quite a bit. Weight will probably be one of the next big considerations in this design.

Is there a reason as to why you decided to have the mechanums on the outside instead of the inside?
Did you guys play around with any other configuration ideas, such as having all 8 of the wheels on the ground at once?

We put the mecanum wheels on the outside because we thought those would be the primary driving wheels, and then the traction wheels would be used in a pushing match or a similar situation. And if a robot with a relatively high center of gravity is moving quickly on mecanum and switching directions, it is much more stable to have the wheels on the outside.

We thought about different things you could do with all 8 wheels on the ground, but it seems like you don't gain a lot with that because you lose both your traction and maneuverability.

Does it spin in place well? It looks like the mecanum wheels might need to be swapped.

We haven't had any problems with the mecanum drive spinning or strafing. I believe our wheels are on in the correct orientation. Unfortunately, we've already made the mistake of putting them in the wrong spot once this year, so we wouldn't want to mess up again :p .

Thanks for all the compliments, and if you have any other questions or comments, please post them.

sdcantrell56

29-06-2010, 03:46

We haven't had any problems with the mecanum drive spinning or strafing. I believe our wheels are on in the correct orientation. Unfortunately, we've already made the mistake of putting them in the wrong spot once this year, so we wouldn't want to mess up again :p .

Thanks for all the compliments, and if you have any other questions or comments, please post them.

The wheels are definitely not in the correct orientation. When viewed from the top the rollers of the 4 wheels should form an X.

Ether

29-06-2010, 08:31

The wheels are definitely not in the correct orientation. When viewed from the top or bottom the rollers of the 4 wheels should form an X.

When viewed from the top they should form an X.

When viewed from the bottom they should form an O.

see this post:
http://www.chiefdelphi.com/forums/showpost.php?p=907174&postcount=18

~

Aren Siekmeier

29-06-2010, 11:49

We put the mecanum wheels on the outside because we thought those would be the primary driving wheels, and then the traction wheels would be used in a pushing match or a similar situation. And if a robot with a relatively high center of gravity is moving quickly on mecanum and switching directions, it is much more stable to have the wheels on the outside.

Also, tank turning with the traction wheels will be easier with a shorter wheel base because you will get less tread rub.

Thelilpluck461

01-07-2010, 12:09

There is only one Solenoid on it, and two accumulators. If you look at the start of the video it is mounted just north of the power switch.

Thelilpluck461

01-07-2010, 21:09

The wheels are definitely not in the correct orientation. When viewed from the top the rollers of the 4 wheels should form an X.

In the pictures on the website the wheels are on the wrong way. If you look at the video the wheels are on the correct way. If you look at the video vs. the pictures the pneumatic and wiring is different, the video is the most up to date version.

CENTURION

16-07-2010, 18:53

Lil' Lavery

16-07-2010, 19:49

Also, tank turning with the traction wheels will be easier with a shorter wheel base because you will get less tread rub.

The inverse is also true, it will also be easier for you to be turned.

A longer wheel base grants more resistance to rotation, both caused by your robot, as well as caused by other robots. This is precisely why 148 and 217 opted to put their traction wheels on the outside, as they only planned to use them in pushing matches and why they were attempting to hold their position.

It all comes down to precisely what you want to accomplish with the drive system. There are reasons and rationales behind all sorts of different wheel configurations.

Aren Siekmeier

18-07-2010, 02:14

The inverse is also true, it will also be easier for you to be turned.

A longer wheel base grants more resistance to rotation, both caused by your robot, as well as caused by other robots.

But the robots motors will have to be back driven for this to happen... The tread rub in tank turning is sideways, but there is still traction inline with the wheels, and this doesn't go away when transverse tread rub is reduced. Rotation of the robot will still require that the wheels turn, and therefore the motors are back driven, or the treads slip. If you are powering the motors, like during a pushing match, you should be fine if the traction is enough.

Lil' Lavery

18-07-2010, 12:48

But the robots motors will have to be back driven for this to happen... The tread rub in tank turning is sideways, but there is still traction inline with the wheels, and this doesn't go away when transverse tread rub is reduced. Rotation of the robot will still require that the wheels turn, and therefore the motors are back driven, or the treads slip. If you are powering the motors, like during a pushing match, you should be fine if the traction is enough.

I bolded the key point in your statement, "or the treads slip." The treads can slip sideways when an external force is applied. Draw up some free body diagrams and think about the problem some more, and it become quite clear why a long wheel base can resist more torque before having the wheels slip.

AdamHeard

18-07-2010, 14:39

I bolded the key point in your statement, "or the treads slip." The treads can slip sideways when an external force is applied. Draw up some free body diagrams and think about the problem some more, and it become quite clear why a long wheel base can resist more torque before having the wheels slip.

Physics are right on this one, that's the beauty of it.

nitbaj

20-07-2010, 16:28

Is the mesh material you used for the eletronics board that
fiberglass-reinforced plastic that is used for catwalks and the like? Or is it just a plastic mesh?

I've been looking around for a simple plastic mesh to use for electronics boards that isnt the more expensive fiberglass stuff, but haven't had any luck.

For the electronics we just used some plastic eggcrate (I think that's what it's called). We just got it at a local hardware store, and it's similar to this:

http://www.goodmart.com/products/1133880.htm

It isn't the sturdiest material, though and after a lot of driving and running into things, the plastic snapped in a few mounting places, so we're planning on swapping it out for something else soon. It is nice for prototyping because it is quick and easy to put on, but I wouldn't necessarily recommend it for a robot you want to be able to take some impact.

CENTURION

23-07-2010, 04:21

Ah, cool...
You might want to try out some of this stuff:
http://www.corrosionfluid.com/fiberglass_grating.aspx#benefits
Or something similar, apparently the FRP (fiberglass-reinforced-plastic) is pretty strong (people can stand on it) though I can't seem to find any with a mesh smaller than 1.5"x1.5", and I'm not too sure about the pricing.

Henry Williams

22-11-2010, 19:41

Just a quick update for everyone interested, since the last post we have fixed the mecanums, we had actually done this a while ago but some of the media was not up to date. We also replaced all the electronics board material with pegboard and we added a shooter so we could use it as a demo/show bot. It is still made of nearly all 80/20 so with the shooter and hopper it probably weighs about the same as a normal FRC bot now (but we haven't weighed it since so thats just an estimate).

Tom Ore

22-11-2010, 20:09

We used these Mecanum's this year. They do look a lot sturdier, but the bolts and rather small and can bend easily. Our Breakaway robot had a similar drive, but without the moveable wheels. We used 4 Mecanums on the inside and 4 traction wheels on the outside, raised slightly. This allowed us to cross the bump easily - the traction wheels were geared down and pulled us over the bump. We could cross the bump easily at full robot speed but when the robot came down on the other side of the bump, the bounce would bend the small Mecanum wheel bolts. The electronics guys didn't like us crossing the bump that fast anyway...

MattC9

22-11-2010, 21:00

Just curious how did you interface the gearboxes to the pods, do you have any CAD files we could look at? Because I was playing around with this idea (like ever one else it seems like) and could not find a good stable way to inter face the transmissions (I was just using the tough box nano in it) also what gear box are you using on it?