Banebot 56mm gearbox - double D related

[RaMoore]

Originally Posted by Joe Johnson

Some questions:
Why do it 3 times?

Did you "temper" the part after the final time you quenched it?

Will you do your test again only with the non-hardened carrier for a baseline?

Thanks for the data.

Joe J.

We did it three times to guarantee that we got hardening all the way to the core of the material. To harden the material further we would have to use higher temperatures and cool it faster and we felt this was unnecessary. Before hardening we were able to easily shave pieces of steel off the edges (using a pocket knife) and after the third quenching we could no longer shave large pieces of material off.

We didn’t temper the material following the third quenching because hardening the material to the point we would need to temper it is beyond our capabilities.

We will repeat the test with one of our unhardened carriers and post results soon.

[Joe Johnson]

Originally Posted by RaMoore

We did it three times to guarantee that we got hardening all the way to the core of the material. To harden the material further we would have to use higher temperatures and cool it faster and we felt this was unnecessary. Before hardening we were able to easily shave pieces of steel off the edges (using a pocket knife) and after the third quenching we could no longer shave large pieces of material off.

We didn’t temper the material following the third quenching because hardening the material to the point we would need to temper it is beyond our capabilities.

We will repeat the test with one of our unhardened carriers and post results soon.

Thanks for reminding me about the pocket knife test. I forgot that bit.

I am not a hardening expert by any stretch. From what I remembered from the one class I had that cover this, I thought that only the last one mattered. But I think I was mistaken. I will think some more about it and try to make a new hardened part tomorrow -- at least one will be using your 3rd time's the charm method.

Joe J.

[Dillon Compton]

Hi Joe,

We are using the 56mm gearboxes with the 16:1 reduction, no chains, one per wheel. I am currently very concerned about this issue, but as you keep re-iterating to do, we are avoiding panic (although I am currently investigating other transmission options, should we decide we need to change our design).

My team has very little machining capability, and no sponsors with the capability to do machining such as manufacturing harder plates for us.

A (sort of) solution that springs to mind is that perhaps Banebots could make the current plates with the "double D" available on their website for sale (preferably at "cost"). Given the ease with which the 56mm planetaries are dis-assembled and re-assembled, having spare plates would potentially function as a "fix" for those teams with a few extra dollars in their budget but did not want to/could not afford to buy a full set of replacement gearboxes. 1/2 way through the competition, one should be able to dismount the transmissions, check for the "bow-tie", and if needed swap the plates.

Just a suggestion - I look forward to an official response/fix/etc... from FIRST and BaneBots, as well as the results of your own, clearly extensive and dedicated, testing and trouble shooting.

Thank you for your continued support with this issue, Joe, and good luck to everyone else who is dealing with this issue.

//Dillon Compton
Team 1394

[RaMoore]

We just repeated our test with the unhardened carriers and we successfully broke our gearbox.

We could clearly notice the damage by the 50th cycle. I'll add pictures as soon as I get a chance. We feel that hardening definitely helped, but whether it is a "fix" more testing will tell.

[Cascade]

So after reading all the posted information and an experiencing an appropriate amount of panic, we are ditching the 56 mm planetary gearboxes for a pair of Andy Marks.

It does not make much sense to us to have worked hard over the last year to build a better organization, recruit more team members, mentors, and sponsors and hope a fundamental piece in the KOP will hold up. Imagine the frustration, tears and anger that would follow if the transmissions fail. Oh and the Rule 7 updated posted today did not make us all warm and fuzzy either!!

Switching was a no brainer.

[Dan Petrovic]

We will be manufacturing many of these carrier plates out of 4140 steel. We are going to purchase 2 foot long stock of it and maybe make as many as we can out of it.

I don't know about drawings for them, but we will create one if we can't find it.

We suspect it will increase the strength about twice as much.

[lips]

I saw this mentioned on another threat in a previous post, has anyone tried welding the two (gear and plate) together? Are there any significant downsides to attempting this, as it seems like a relatively simple (if welding is available) fix...

Also, has anyone tried pinning the two? That idea was tossed around this evening, milling and pinning at three or four points the gear to the plate. This would support the central pivot on the double D, if nothing else.


Just some thoughts, drawing and (hopefully good) results forthcomming

[Kevin Sevcik]

About hardening, I don't think the 3 times repeating will get you a lot of extra depth, but I'm going off my one class a couple years ago, same as Dr. Joe. If you're not heating it high enough to transform the metal and break your previous hardening, then your hardening isn't going to work. If you are heating it hot enough, then you're breaking down the previous hardening. I suspect what you might be doing is a poor man's version of carburizing, building up a layer of carbon from the volatilized motor oil, then letting it soak in by heating up the part again.

At any rate, some brief googling gives me a little better info for those trying the homebrew hardening. The austenizing point is apparently slightly past where the metal stops being magnetic. So heat a bit past where a magnet won't stick to it anymore. I'm also curious if the carrier can withstand a water quench if anyone has a carrier they wouldn't mind possibly warping or cracking.

[Joe Johnson]

Originally Posted by lips

I saw this mentioned on another threat in a previous post, has anyone tried welding the two (gear and plate) together? Are there any significant downsides to attempting this, as it seems like a relatively simple (if welding is available) fix...

Also, has anyone tried pinning the two? That idea was tossed around this evening, milling and pinning at three or four points the gear to the plate. This would support the central pivot on the double D, if nothing else.


Just some thoughts, drawing and (hopefully good) results forthcomming

Good suggestions.

My thought on welding is that it will be hard to do without damage to the bearings and if the bearing survive the heat, I am not sure that the joint will be that much stronger -- the problem is the base metal gets pushed out of the way.

Pins are another thought. It may be that a pin in the middle of the rounds could take more torque without taking away more than they give. I think it would take an FEA to be sure as the loading gets pretty complex for pen and paper guys like me. Anyone up for an FEA optimization problem? How many pins and where to locate them to reduce the stress by a factor of 2?

PM me and I will coordinate efforts.

Thanks. Keep the ideas coming.

Joe J.

[Joe Johnson]

Originally Posted by RaMoore

We just repeated our test with the unhardened carriers and we successfully broke our gearbox.

We could clearly notice the damage by the 50th cycle. I'll add pictures as soon as I get a chance. We feel that hardening definitely helped, but whether it is a "fix" more testing will tell.

This is pretty great news. Now we have a clear this fails / this does not case.

Well Done. I had forgotten to tell RaMoore when I asked them to do this that we will get them new parts to replace the damaged ones.

As to pictures. Yes, we need them.

What we need now is a mat'l lab that can reverse engineer the mat'l of the shaft and of the carrier plate ASAP. Does anyone know of a lab that will do this quick and free?

If we know the material it may be possible to design a heat treating process that can give us the harness we need. If that is possible, then one possible path toward a solution is to locate a heat treat source that will set up a carrier exchange program.

But, first things first, anyone Got Lab?

Joe J.

[Lil' Lavery]

I read in one of the close threads (please forgive me for being lazy and not finding the specific post) that when a team tried to pin the two pieces together, they actually broke their drill bits on the output shaft. I'm not sure what bits they were using specifically (as they didn't mention), so this may or may not be problem in terms of teams doing this themselves.

[ZZII 527]

Originally Posted by Lil' Lavery

I read in one of the close threads (please forgive me for being lazy and not finding the specific post) that when a team tried to pin the two pieces together, they actually broke their drill bits on the output shaft. I'm not sure what bits they were using specifically (as they didn't mention), so this may or may not be problem in terms of teams doing this themselves.

That was us. It was a very makeshift job with ordinary 1/16" bits. The carrier drilled fine, but trying to go through the harder shaft, they broke. But other teams with better bits might have good results. Also, I think someone mentioned milling slots for the pins instead of drilling straight through.

[MrForbes]

The shaft is probably hard enough that it would need to be ground (or milled with a carbide cutter) to make the slots in it for pins.

Pins will act kind of like splines...the angle is better than the double D configuration. Perhaps spreading the load out among two pins, the double D, and fixing any end play problems, would be enough to make it work?

although the end play stuff gets more disconcerting the more I look at it, when I reassembled the transmission that I took apart for the exploded view this morning, it had somewhere around .030" play between the bearing and the snap ring on the shaft. There is a lot of slop there, the bearings are a press fit on the shaft, and it may be necessary to also press the inner bearing onto the shaft fully while supporting the outer bearing, and then install a selective washer between the inner bearing and the shaft snap ring.

This is kind of like working on car transmissions....which I've done a bit of.

[RaMoore]

Pictures of unhardened carrier as promised.

The damage is not as extensive as other teams have seen, but a clear difference from the hardened one.

As for the repeated cycles of heat, we tested with the knife scratch and it did seem to change it, for whatever the reason. I don't think any of us has had a class on the subject just some practical experience and some googling. I think we could try Kevin Sevcik suggested method for hardening if we can get it that hot.

[dtengineering]

We have four 56mm BB trannies running our mecanum drive. The 8" AM mecanums are directly mounted to the BB drive shaft which is supported on the outboard end quite securely. (photos at http://www.chiefdelphi.com/forums/sh...highlight=1346 )We "broke in" the gearboxes for at least 30 minutes of continuous operation before installing them.

We have been running our robot for... well... less than 30 minutes and noticed one or two of the trannies were beginning to show significant backlash.

We took the transmission apart and noticed that while the final drive plate had not failed completly it was very close to failure as discussed in this thread.

This is after very low use with ONE CIM in a relatively low dynamic loading (no chain to snap) environment.

We have noticed that the two BB's that came in the KOP have less backlash. They also fit much more nicely onto the CIMs than the two we bought in early January. Perhaps there is variation from batch to batch to deal with as well.

In any case, here is one more report of BB 56mm failing (or at least coming close enough to failure to count) under ONE CIM load... and at extremely low cycles, too.

We are also pondering solutions, but it seems manufacturing a new plate, or having one manufactured is what we are going to have to do if we want to use these trannies.

Jason