pic: One Tough Train

[Rod]

loading photo ...

[sanddrag]

Rod,
You don't come around these boards too often but when you do I am certainly impressed with your work. You are becoming a great robot designer. Have any of your designs been built? If so, I'd like to see some pictures.

A few notes:
-It might be better to mount the CIMs by the 10-32 threaded holes in their face rather than claming them on the can.
-Why use a coupling on the CIM shafts rather than just putting a gear?
-The transmission directly drives the wheel, I LOVE IT!
-The transmission has three plates, make sure there are never 3 bearings/bushings on a single shaft.
-For the sprockets, what size are they? I would recommend going the smalles you can (which would be 13T for #25 with 1/2" bore)
-Finally, have you considered a welded box tubing frame, they can come out to be very lightweight.

Last, I would like to offer you a picture of our frame in case you wanted to take any design cues from it. http://www.chiefdelphi.com/forums/pi...le&picid=10020 The pillow blocks slide for chain tension. Big tubing is 1.5x2 (probably should have used 1x2 and the small tubing is 1x1. All 1.8" wall.

Again, NICE WORK!

[Tristan Lall]

That looks interesting....

Originally Posted by sanddrag

A few notes:
-It might be better to mount the CIMs by the 10-32 threaded holes in their face rather than claming them on the can.
-Why use a coupling on the CIM shafts rather than just putting a gear?
-The transmission directly drives the wheel, I LOVE IT!
-The transmission has three plates, make sure there are never 3 bearings/bushings on a single shaft.
-For the sprockets, what size are they? I would recommend going the smalles you can (which would be 13T for #25 with 1/2" bore)
-Finally, have you considered a welded box tubing frame, they can come out to be very lightweight.

I mostly concur with sanddrag, except for that sprocket thing.

A couple of years ago, in 2003, Woburn used those same little sprockets, and promptly broke something like 13 #25 chains over the course of three events. Since there's no such thing as too much horsepower, the chain simply wasn't up to what we demanded of it. (It undergoes higher stresses as sprocket radius decreases.) You might consider looking at Tsubaki's General Catalog, Section 1, pages A-6 and A-7, where it gives power vs. speed vs. teeth-on-small-sprocket charts for #25 and #35 chains. (I'm always recommending this, it seems.)

Some quick calculations say that for your 5.5 in wheels at 6.5 ft/s, you need a wheel speed of 271 rpm. Plug that into the chart, and you get a rated power of 0.17 HP for #25. Two CIMs at 341 W each are good for about 0.91 HP, or 5.3 times your chain's rated capacity. (You're screwed ; that's even worse than Blizzard 4, in power-to-rating terms--it was an already-horrifying 5.1, even after accounting for its extra power!)

Of course, there's an answer: move up to #35, where the rated power for a 13 tooth sprocket (note that it's bigger than the #25 version, in all dimensions) is 0.94 HP, which is perfect.

I don't know how much of a safety factor chain manufacturers build into these ratings. I've got empirical evidence to demonstrate that it's not 5, however. Maybe 2, if you like to live dangerously--but gambling that the safety factor exceeds 5.3 really won't be pretty.

[sanddrag]

I'm not sure how or why you broke your chains but we ran a 6 motor 2 speed with #25 13T sprockets without a problem. While you are correct in saying you can easily exceed the working load, the breaking point is over 1000lbs. Also, it is not about the force the motors produce, it is about the force the wheels can hold on the floor. But yes, larger is safer.

[Tristan Lall]

Originally Posted by sanddrag

I'm not sure how or why you broke your chains but we ran a 6 motor 2 speed with #25 13T sprockets without a problem. While you are correct in saying you can easily exceed the working load, the breaking point is over 1000lbs. Also, it is not about the force the motors produce, it is about the force the wheels can hold on the floor. But yes, larger is safer.

I think a lot of it had to do with the style of play in 2003, versus this year. You mentioned elsewhere that your 2005 robot didn't have much pushing and shoving to do; by contrast, there were plenty of abrupt starts and stops in the 2003 game, not to mention outright collisions, where closing speeds could top 20 ft/s. Those represent shock loads not accounted-for in the 1050 lbf static breaking load (actually, for cheap #25 chain, it can be as low as 780 lbf, per the ANSI spec.). Add to this the fact that the three-motor transmissions would accelerate the robot very quickly, and reverse direction just as abruptly, and it's clear that the chains were doomed. But even with a much more relaxed driving style, and better chain tensioners than we had in 2003, 5.3 times the working load doesn't seem too safe.

[Conor Ryan]

Originally Posted by Tristan Lall

That looks interesting....

I mostly concur with sanddrag, except for that sprocket thing.

A couple of years ago, in 2003, Woburn used those same little sprockets, and promptly broke something like 13 #25 chains over the course of three events. Since there's no such thing as too much horsepower, the chain simply wasn't up to what we demanded of it. (It undergoes higher stresses as sprocket radius decreases.) ....

can't this all simply be solved by replacing the chain system with timing belts? Cars use them and put them under alot more stress than we ever will, no to mention are quite a bit litter than chains.

[sanddrag]

Originally Posted by cdr1122334455

can't this all simply be solved by replacing the chain system with timing belts?

No, If you break a chain, you are sure as heck going to break a timing belt, unless it slips first which would be almost as bad.

[Rod]

This is our drive train this year, it was designed in preseason and modified for this year's game. And yes we always use #35 chain in our drive trains.

[Rod]

Originally Posted by sanddrag

Rod,
You don't come around these boards too often but when you do I am certainly impressed with your work. You are becoming a great robot designer. Have any of your designs been built? If so, I'd like to see some pictures.

A few notes:
-It might be better to mount the CIMs by the 10-32 threaded holes in their face rather than claming them on the can.
-Why use a coupling on the CIM shafts rather than just putting a gear?
-The transmission directly drives the wheel, I LOVE IT!
-The transmission has three plates, make sure there are never 3 bearings/bushings on a single shaft.
-For the sprockets, what size are they? I would recommend going the smalles you can (which would be 13T for #25 with 1/2" bore)
-Finally, have you considered a welded box tubing frame, they can come out to be very lightweight.

Last, I would like to offer you a picture of our frame in case you wanted to take any design cues from it. http://www.chiefdelphi.com/forums/pi...le&picid=10020 The pillow blocks slide for chain tension. Big tubing is 1.5x2 (probably should have used 1x2 and the small tubing is 1x1. All 1.8" wall.

Again, NICE WORK!

Thanks
A few answers
The gears used are a 55 tooth 3/8 face 20 DP for output and a 5 tooth stem pinion on the end of the Cim motor. The stem pinion has a 3/8" Dia. so it can not be mounted on the Cim shaft. That is why the Cims use a coupling and are not face mounted. I have put 3 bearings on a shaft many times and never had a problem. Our robot this year uses the same gearboxes and works perfect. The sprockets are 14 tooth #35 with 5/8" Bore. This year we had to add weight to our robot (17 Lbs.) We have a 55 Lb. drive train this year, that does not count the arm. So next year if we have a 68 Lb. drive train it should be about right. I read these boards every day, just don't post too much. Our designs have been built, they are posted here both the CAD drawings and this year's Robot. CAD drawings are titled Light weight drive train, we built a slightly modified version of it.
See Ya

[sanddrag]

For 3 bearings on a shaft, it is a geometric overconstraint. Unless you can keep tolerances to under .001, you are losing efficiency.

[Rod]

We use plain sleeve flanged bearings. The center bearing is bored out .020" over and is more of a gear spacer than a bearing.

[sanddrag]

Ah, okay. So you really don't have 3 (close fitting) bearings on the shaft. Also, now I can see the need for couplings. Looks very small/compact!

I'm not sure if this would be a good idea but it seems like those couplings are taking up a lot of space in your otherwise compact design. If you got a gear just big enough to bore out and key to fit over the CIM shaft to keep the tooth count small (for 20P I believe this would be a 12T) then you could increase the 55 tooth gear just a little bit (maybe to 60) or maybe even keep it that size for a faster robot then you could move the CIMs in closer and mount them by the face to that first red plate.

[Rod]

Originally Posted by sanddrag

Ah, okay. So you really don't have 3 (close fitting) bearings on the shaft. Also, now I can see the need for couplings. Looks very small/compact!

I'm not sure if this would be a good idea but it seems like those couplings are taking up a lot of space in your otherwise compact design. If you got a gear just big enough to bore out and key to fit over the CIM shaft to keep the tooth count small (for 20P I believe this would be a 12T) then you could increase the 55 tooth gear just a little bit (maybe to 60) or maybe even keep it that size for a faster robot then you could move the CIMs in closer and mount them by the face to that first red plate.

We are at 11:1 now. If we used 12T to a 60T we would be at 5:1> That would be way to fast for us with no low end.

[Kevin Sevcik]

How stiff is that chassis, anyways? It definitely looks good, but we tried a bent sheet chassis one year and we definitely had problems with it flexing. and especially with it twisting since it didn't have any stiffness at the corners.

[Rod]

Originally Posted by Kevin Sevcik

How stiff is that chassis, anyways? It definitely looks good, but we tried a bent sheet chassis one year and we definitely had problems with it flexing. and especially with it twisting since it didn't have any stiffness at the corners.

The bottom plate is 1/4" thick Alum. I hope it will not flex too much