23-06-2005 00:59

sanddrag

Re: pic: Planetary...Wheel?

Can we get a view without that outermost plate and bearings so we can see the gears inside?

23-06-2005 01:10

Re: pic: Planetary...Wheel?

Is a 10 inch wheel too big? (I don't know, so I ask), and as big, I mean overly large for the size of a robot that would be made for a FIRST competition....

Also- would it be possible to make this a 2 speed?

23-06-2005 01:30

Matt Krass

Re: pic: Planetary...Wheel?

Sanddrag, I'll post an exploded view tomorrow showing the planetary gearset.

Craig, my team has used 10" wheels with much success since the 2002 season. Also, I suppose you could replace the first stage with a shifter but that would defeat the point somewhat, this is meant to be simple and compact.

23-06-2005 03:28

team222badbrad

Re: pic: Planetary...Wheel?

I have been thinking about an in wheel transmission/gearbox like this, but the thing that worries me most is impacts and side loads.

I suppose if you can find a solution to that it would be great.

Keep the ideas flowing.

23-06-2005 11:51

Andy Brockway

Re: pic: Planetary...Wheel?

Matt,

Is the ring gear attached to the wheel? If so, what forces the wheel to turn? Most planetaries will allow the moons to spin around the sun when the outer ring is stationary.

I suppose this will be cleared up when you post the exploded view. I believe you want the moons attached to the wheel with the outer ring fixed to the frame.

23-06-2005 13:04

Jeff Waegelin

Re: pic: Planetary...Wheel?

As I understand it, having seen the exploded view, the sun gear is on the motor output, the planets are stationary, and the ring gear is locked to the wheel. If I'm not mistaken, this setup will cause the ring gear to spin in the opposite direction as the motor, with a reduction equal to the number of teeth on the ring, divided by the number of teeth on the sun gear. Is this correct?

Over all, I was pretty impressed with this design. Matt showed it to me a few weeks ago, and I was considering using it in a summer project, though I ended up going in a slightly different direction. It seemed to me like a pretty easy, compact way to do a direct-drive gear reduction.

23-06-2005 13:09

Matt Krass

Re: pic: Planetary...Wheel?

It's not shown in the picture (yet) but six bolts will hold the ring gear to wheel, it's actually 1/8" thick, while the wheel has an 1/8" flange, total wheel thickness 1/4". The Sun and Moon gears are all idling on shafts (see those bearings in the plate? those are for the Sun and four Moon gears) while the ring will spin, and in turn the wheel. I will post the exploded later today, for now I need to clean my room before my parents yell .

23-06-2005 15:40

Matt Krass

Re: pic: Planetary...Wheel?

I just posted the exploded views, two are already approved, the other two should be shortly, I hope this clears up any questions.

23-06-2005 16:24

Andy Brockway

Re: pic: Planetary...Wheel?

This is the same as I see it. I missed the planets being fixed. Do you know what the ratio would be if the outer ring is fixed and the wheel is attached to the planets?

A very interesting design. The first of the summer season!

23-06-2005 16:48

RogerR

Re: pic: Planetary...Wheel?

are you sure about this gear thickness? are the planet and sun gears an equal thickness? and what DP are these gears? while i don't have any calculations (i'm in the library "studying" for a final), 1/8" seems a little thin to me.

23-06-2005 17:08

Not2B

Re: pic: Planetary...Wheel?

Very cool. That could be a very interesting design, not only for drive wheels, but for other devices.

Here's my question - What keeps the ring gear from moving cross-vehicle? I think there are a couple of ways to solve this, but I was wondering what you had in mind.

23-06-2005 18:38

Tytus Gerrish

Re: pic: Planetary...Wheel?

what's going to support the side load of the wheel? . looks like all the weight of the robot is going to be forced onto the bottom planet gear and In all my experience gears don't like to be smashed together like that at all. , also if that wheel takes any side load it's going to come off. both of those things can be solved with a big bearing. but Where Re you going to get one? those big bearings are hard to find, not to mention expensive. Thats a cool design but if i was planing on making a robot i would use something much simpler and less expensive.

23-06-2005 19:35

Matt Krass

Re: pic: Planetary...Wheel?

The side load problem is the one next on my list, this is a completely new area for me, so I'm kind of winging it, I have a few ideas to prevent the carrier gear from slipping off, I'm not sure about the downward forces on the planets, I was thinking of rotating them 45 degrees to spread the load across the two bottom gears, this is all brand new territory for me, and as far as I know, FIRST in general, though I could be wrong about that. Does anyone know of any similar designs used in FIRST?

23-06-2005 22:15

Kevin Sevcik

Re: pic: Planetary...Wheel?

I'm not an expert on gears by any means, but I think just spreading the load across the gears won't help that much. Gears are meant to operated at fixed distances on fixed shafts. 25 pounds of force pushing the gears into each other isn't going to make them happy. Even if you could get it down to 5 pounds, you'd still lose a lot of efficiency. I think your best bet is to figure some way to get a bearing supporting the wheel so you're not using the gears as a bearing sort of surface. This would help with the side loads too.

I'm thinking a big thin section ball bearing that goes around your array of bearings there and slips inside a rim on the wheel. At that diameter a brass bushing might work as well.

24-06-2005 00:57

Jizvonius

Re: pic: Planetary...Wheel?

Hey nice design. One way to reduce the range of loading on the gears would be to find a way to support the wheel on it's own bearing. It may have to be a needle roller bearing with a large inner diameter to keep the design compact but it should work out with a bit more structure. That way the wheel loads go to the bearing and the gears don't have large time varying loads that will occur even without impact with the present design.

Also, you can get away with using fewer gears now that the loads don't have to be distributed that much. That should save a little on weight.

24-06-2005 01:02

Matt Krass

Re: pic: Planetary...Wheel?

Thanks for all suggestions, I'm taking them all under consideration, now I have a question for all you gearheads out there...

What the heck is a needlepoint bearing?

I'm learning, one step at a time..

24-06-2005 01:11

RogerR

Re: pic: Planetary...Wheel?

needle bearings use thin cylindrical rollers instead of a balls (in the case of ball bearings).

24-06-2005 01:27

Jizvonius

Re: pic: Planetary...Wheel?

Indeed, and for larger diameter shafts, they are generally a whole lot smaller than ball bearings that will get really huge and heavy.

Andy Baker's suggestion in the other thread was also a really good idea for taking the load off the gears and constraining the wheel motion, just depends on what you choose.

Jevawn

25-06-2005 14:03

dlavery

Re: pic: Planetary...Wheel?

Matt -

Look here for some examples of needle bearings. You can get them straight from the manufacturers (sometimes - depending on how many you want to buy), or from standard sources like MSC, McMaster-Carr (see page 1023 of their catalog), Small Parts, etc.

-dave

26-06-2005 13:06

Joe Johnson

Re: pic: Planetary...Wheel?

What is the purpose of the design?

It is fairly clever but without knowing what you are planning it is hard to be fair.

In generally 3 things worry me.

Others have said it already but imho, you need to get independent bearings on the wheel/sun. Gears are not intended to be bearings. You have little chance of success if you depend on the gears for support.

The thinness of the wheel. I think you will want to consider thickening up the wheel. Wheels take considerable side loads, more than I would expect a 10" dia, 1/4th thick spoke splate to be able to support.


Finally, gear tooth face width. This is 2 problems in one. I think 1/8th thick gears will be hard to keep aligned.

Without actually counting teeth it looks like you've got about a 15:1 overall reduction from the motor to the wheel. But more importantly, 2 Chiaphua motors put out a lot of torque, more than those gears look like they can hold if my gut is right.

Without actually doing the calculations to see if they are going to fail, you can get a feel for things by looking at designs you know work.

Without actually counting teeth it looks like you've got about a 15:1 overall reduction from the motor to the wheel. Thinks about it this way: Look at the face width of the gears inside a planetary gearbox. Take the Dewalt for example. Look at the 2nd stage. See how thick those gears are? Well, your ratio about the same as that second stage BUT BUT BUT... ...2 Chiaphua motors will put out 5-10 times as much torque as a single drill motor. So... you've got many times the torque trying to be supported by gears many times thinner. Something has got to give.

Just some things to think about.

Joe J.

26-06-2005 16:11

Matt Krass

Re: pic: Planetary...Wheel?

I appreciate your comments, as well as everyone else's, this design was a mere idea I came up with one day, I didn't think it anywhere near fully through and I'm not sure if it will ever be feasible. That being said, I'm currently sketching ideas to get independent needlepoint bearing support for the wheel, hopefully I can find a solution to the side load and gear load problems.

Also, The wheel thickness in this diagram is fairly thin I admit, but as stated before this is more of a conceptual model than anything to build from, same goes with the thin gears, I'm sure I could find better ones up to the job, and I will when I finish "rev 2" the "Production model" over this summer. I'm fairly busy, just starting work and handling my last summer in high school and all, so it may take some time to finish. Also, I've devoted some of my low amounts of CAD time to version 3.0 of my Omniwheel, which I'm posting right after this, based on suggestions offered in that thread to improve the design, it was a little easier and less demanding, so I figured I might as well get it out of the way. I appreciate everyone's help with this and I hope a final, workable design can be realized, and that it may be of help to a team at some point.

And now back to the wizard of Cogs...

26-06-2005 21:55

sanddrag

Re: pic: Planetary...Wheel?

If this turns out to be more trouble than it's worth (which I'm not saying it necessarily is) there is always the option of making a traditional fixed ring planetary, and having the output of that be your wheel axle. You'd have to take some precautions so that wheel impact doesn't happen or if it happens that it doesn't affect the gears. It would end up being a little bit long to space out the bearings (probably an inch of spacing between them will do), but you could have the wheel double back over the gearbox sort of like the MER rovers. You could have a deep dish wheel so the gearbox goes inside of it a bit.

I'm even thinking maybe a Chiaphua/Dewalt directly driving a wheel by modifying the output shaft and end of gearbox to give better support for holding a wheel. The only thing is though you wouldn't want to go any larger than about a 5" wheel straight on a Dewalt, but at least I think that is enough to clear it so the gearbox can go inside the wheel a little bit.

The thing with making planetaries is that I haven't been able to find a source of wide face internal gears. The largest I've found is 5/16 and it is brass.

27-06-2005 16:05

Jizvonius

Re: pic: Planetary...Wheel?

I think the internal gear idea can work if the pitch diameter is large enough. Then it wouldnt require nearly as much force on the gear teeth as it would be applied far from the center of rotation. This thread actually made me resurrect an old idea that can possibly make a very compact drive module. The gears(except the internal gear), brackets and shafts are missing from the inventor drawing, but the spacing has been worked out in 2D. The model so far should give you an idea of what I'm working towards though. The gear ratio ends up being about 15:1. The weight looks like it will end up being about 8 lbs and the size is about 8"x6"x6". The design obviously needs more thought, but I figured that I'd put it out there.

I have many variations of this design that move components around for different purposes. This one is to minimize the amount of gearing and space for the motor-gearbox-wheel system.

Jevawn

BTW i did this last night at around 4:30am so any resulting crappyness of the model can be attributed to that.