Reverse Gear

[FizMan]

I'm trying to learn and develop some ideas of mine. Unfortunately, my background in this type of mechanical engineering is somewhat lacking.

Now I'm trying to develop a way to have a wheel go in reverse withOUT having the motors reverse. That is, my goal is to have a single gearbox with all the motors in it, and implement a reversing gear for the wheels I want.

Is there another method to accomplish this besides a reverse gear?

Here's how I concocted my reverse gear (on the toilet no less); so I also need to know whether or not they'll be any change in torque/speed between the two gear settings. I used a pneumatic cylinder to push one gear out and put two gears in (changing the direction). The Cyan Gear on the left is the Input gear (say from the gear box) and the Cyan Gear on the right is the output gear (say to transfer power to the wheel). The Red Gear is one direction, while the Purple Gears provide the reverse direction.



-forwards-



-backwards-




('course, the "forward" and "backward" direction is completely arbitrary at this point... so if this is implemented, the "backward" gear in this example may turn out to become the forward gear)


See any problems with this setup? Any comments are much appreciated!

[Kevin Sevcik]

This should work pretty much as you plan, although obviously the shafts need to be supported on both side of the gear and such. The speeds won't change in theory, as the gears you're shifting are just acting as idlers. In reality, you'll get a drop in efficiency when you switch from the large gear to the two small gears, which will affect your speed slightly.

[Pat Roche]

Looking at the only other problem I can see besides the one previously mentioned is that when you make that shift between forward and reverse you might not mesh your gears. You may want to create a way so that the gears slip better...possibly a taper or something...

If you have any questions about what Im saying please ask...(lack of sleep bogging my mind down)

-Pat

[Kevin Sevcik]

hrm. I have no way of drawing this at the moment, but here's a large improvement on your design.... you have the large and small idler gears permanently engaged, and you shift that last gear back and forth to make the shift. to do that, the last gear should be an idler as well, and engaged with a really long face width gear on the other side that it never disengages from. or it should be turning on the same shaft with a really long gear that won't disengaged from what it's turning. This way, you only have to move one gear. now, that probably doesn't make much sense, so I'll try to get a drawing together later today.

[JVN]

Originally Posted by FizMan

Any comments are much appreciated!

In what program did you make these animations? It looks really cool, and could be useful for sharing ideas beyond the standard napkin sketch.

Thanks,
John

[greencactus3]

actually, having two gears mesh and having a chain over 2 sprockets and having a dog run between the two sets will work too.
anyone understand??
so...

g...sp
e...ro
a...ck
r...et

M...|
E...C
S...H
H...A
I...I
N...N
G...|

g...sp
e...ro
a...ck......."dogrun" between these two(gear/sprocket)
r...et

[FizMan]

First off, thanks for the responses!

The program I used is just your run-of-the-mill 3D animator things. Cinema 4D XL (http://maxon.net) One of the splines in it is "cogwheel"; let's you edit things like the number of teeth; size of the teeth, bevels, radius, etc.


I couldn't quite visualize what you were getting across there Kevin... I look forward to your sketch Also, you mentioned that I'd lose efficiency with using the two smaller gears over the larger one; what kind of change in speed would you guestimate I'd lose over that? 5%? 1%? More? Less?

And Greencactus... I think I see where you're going with that... I went and modelled a full-size reverse-gearbox for all four wheels based on that. Unfortunately, my ISP is having techincal problems right now and I can't access many sites (including my web-host) so I can't upload the animations. But after putting them together, it's going to be iffy whether or not this system is feasible for FIRST. The only plus side of this is that it'd only need one set of motors and one gearbox and one transmission for a four wheel steering system... so I don't know; unless we can find a way to cut down the weight of this system... crunch some numbers; maybe it won't weigh toooo much... When I upload the animations, you'll see I already tried to squeeze some weight out of this by mounting all the shafts on the same bases and whatnot.

I suppose my next question is, what's the easiest (and light-weight) way to shift the dogs in this situation? Grrrr I wish I could upload the animations...

EDIT: I just realized I can put attachements here, so I'll go and convert some .gif's and upload them in a minute.

EDIT: Okay, I added the two of animations; but I had to reduce the quality to bring down the file size so that the ChiefDelphi forums would accept them. So I apologize for that. Anyways, first off, I wireframed the base boards, but they didn't come out too well when I dropped the quality, so you may or may not notice a few grey lines, that's the bases. Don't worry aboot them. Also, I forgot that the holes in the right gears (for the dog to fit into) don't move enough to fit the animation, so you'll see them jump (ignore it )

The power input is on the left with just the blue/cyan sprockets. The output is on the right. The two cyan gears on either side connect via a chain (didn't model the chain ), and the two blue gears connect via the intermediate purple gears. The red gears are actually the dogs that slides between the Blue & Cyan gears.

With the exception of the Blue/Cyan gears on the far left shaft, ALL of the rest are NOT connected to the shafts; they're all free running.

I figure you could run a chain from the dogs to their respective wheels.


So with this setup, you would move the individual dogs and output reverse/forwards to the different wheels, and have them ALL powered by the SAME set of motors/gearbox/transmission; eliminating the need to have two of each or to split up the power.

But yeah, seems a bit extravagant and heavy... but with an all-wheel stearing system; my logic tells me that you'd need to be able to reverse/forwards individual wheels as when they rotate, you'd need to choose which ones to reverse in order to turn the robot. And I don't like the idea of putting together four individual transmissions, gearboxes, and motor mounts for four different wheels...

Thoughts?


EDIT: My ISP is working again so I uploaded the better quality animations:






EDIT:
In terms of saving weight... I take it plastic gears aren't really an option eh?

[Max Lobovsky]

rather than using both sprockets and gears for the reverse system like green suggested, you could just use two different sized gears so that the smaller one has room for a middle gear while the larger one just meshes directly with another larger output gear.

small gear----------------------large gear
intermediate small gear [A]
small gear ---shifting dog-------large gear

note that [A] will be occupied by the large gears also because they are, well, larger than the small gears

and as for plastic gears, they won't in reality save weight because the weight of a plastic gear for the power it can convey is actually higher (unless maybe of some high performance stuff that you probably wont find a gear made out of) than the weight-to-power ratio for steel gears. I'm pretty sure even brass gears (i think most small non-plastic gears are brass) will have a better ratio. [edit] i realized how dumb and not scientific my "strength to weight ratio" sounds. I'm not sure which property of the material is the "strength" in this case, but basically, by strength i mean its resistance to shear which is probably the primary method of breaking a gear, shearing off its teeth. [/edit]

[RogerR]

Originally Posted by maxlobovsky

... I'm pretty sure even brass gears (i think most small non-plastic gears are brass) will have a better ratio.

actually, brass is denser than steel, in addition to being weaker.

[Max Lobovsky]

Originally Posted by RogerR

actually, brass is denser than steel, in addition to being weaker.

right, i understand that, but i thikn it may still have a better ratio than that of nylon or whatever the gears are made out of.

[RogerR]

Originally Posted by maxlobovsky

right, i understand that, but i thikn it may still have a better ratio than that of nylon or whatever the gears are made out of.

if you have the weight for a brass gear, then you might as well use a steel one because, as i stated above:

a) steel is less dense than brass (therefore lighter)

b) steel is stronger than brass

[Cory]

This is a cool idea and all, but it's a whole hell of a lot heavier and more work and possibly complicated, than making 4 individual gearboxes, or two gearboxes and using bevel gears to transfer power to each wheel.

Roger- I think he means if you only have a choice between brass or plastic. Often small gears with a small face witdth only come in brass or plastic.


Cory

[RogerR]

Originally Posted by Cory

... I think he means if you only have a choice between brass or plastic. Often small gears with a small face witdth only come in brass or plastic...

actually, PIC gears go to as small as a 1.58 mm face, with an OD of 6.5 mm (not that i'd personally use anything that finely pitched on my drivetrains)

also, depending on the type of brass/acetal used, the plastic could end up being stronger (i'm not sure which type of brass is used in gear making).

on an unrelated note, i found a pretty good site for looking up this sort of stuff:
matweb