Banebot 56mm gearbox - double D - RELOADED

Technical Motors
1

This thread is actually a continuation of this thread.

The thread was getting so long that I was worried that folks with dial up would miss the ship date waiting for the page to load.
It is a complex issue so I will try to summarize.

History:

  • FIRST sourced Banebots for the 2007 KOP transmission in the Summer of 2006 after having evaluated prototype transmissions in both the 1-CIM and 2-CIM configurations.

  • Shortly after kickoff, I became concerned about the double D joint.

  • Analysis and Bench tests confirmed that there was reason for concern.

  • Earlier this week a team first reported excessive wear in the D joint (the bow tie effect) confirming the failure predicted by analysis and demonstrated by bench tests.

  • More test were done and a few more reports of bow tied carrier plates.

  • FIRST published the issue in UPDATE 5, warning teams that there is an issue and that folks were working on it.

  • Much help has been offered by the FIRST community to address this problem.
    The data is not 100% certain, but here is where we stand as best as I can summarize:

  • In some (but not all) conditions, the double D hole in the carrier plate will plastically deform into bow tie.

  • The field failures up to this point show that the bow tie opens up gradually (many cycles not several fatal blows). As it fails, more and more backlash is observed in the gearbox.

  • It is reasonable to expect that bow tie will eventually (after 100’s of cycles) resulting in complete failure (though as of this moment, complete failure has not reported except as a result of a torque to failure bench test I performed)

  • The carrier plate is significantly softer than the shaft.

  • While the exact alloy of the carrier plate is not known, it is believed that the composition of the plate does not lend itself to easy hardening.
    What is next:
    Couple of teams have volunteered to be best subjects.
    They will run a baseline set of tests to determine the extent of the problem
    These teams will act as our proving ground to evaluate proposed solutions

Working together with Banebots, FIRST, and the ChiefDelphi.com I believe we are closing in on a set of solutions to this issue. Based on my communications with Ed Yackey from Banebots, folks from FIRST, and many in the larger FIRST community, I can assure you that everyone is committed to working this issue to a successful conclusion.

So what is that solution?
I cannot say 100% for sure at this point because there are still a few unknowns that could change the answer, but this is what I believe is the most likely scenario at this point.

**Plan A: **New harder (RC 23) carrier plates. Hardness equal to the hardness of the output shaft. This should result in close to a 2X increase in the max torque

Plan B: New harder carrier plates (RC 23) with square hole (rather than round with double D) Plus modifying or making new output shafts with square drive. Should give approx. 2X increase in max torque over Plan A.

**Plan C: **Same as plan B only with harder (RC 40) carrier plate and output shafts. Should give approx. 1.5X over Plan B **1-CIM motor:**It is my opinion that Plan A will address this issue

**2-CIM motor:**It is my opinion that Plan A will possibly address most of these failures but in order to fully address this issue for teams that are using 2-CIMs, we may have to go to Plan B. Status:
We are working to have a sample solutions of Plan A to test for the weekend. We are in the process of planning Plans B & C but believe me, we will have them ready when the time comes. Also, I have not listed a Plan D but trust me, we will come up with a Plan D and implement it if that is what it takes to make everything come out in the end.

Closing:
This whole thing stinks. Nobody wanted this. But, I am proud of the many people who were part of the solution. And, yes, I say solution. The proof of the pudding is in the eating but I am quite confident that we have all the ingredients of a tasty dessert on the table.

Best Regards,
Joe J.

UPDATE:
I have just attached prints of the carrier (4:1 & 3:1) and the output shaft. I have versions with a 11/32" square hole. This is a commonly available broach size that is just under the 9mm flats on the DD (11/32" = 8.73mm).

I put these prints up here for 2 reasons:

[LIST]

  • 1st to get more eyes to look at the prints to find dimensions we have missed or mistyped.
  • 2nd to allow teams with appropriate resources to make the parts for themselves (and hopefully report back their success or failure)
    [/LIST][INDENT]Before I start a panic, note that NONE of the plans above involve teams making these things for themselves. If a team CAN make them for themselves, they may do so, but rest assured that we will not leave those who cannot stranded.
    [/INDENT]
    Some notes:
    The material is spec’d at 1040 hardened, tempered to RC40. That is harder than we need. There are some who feel that the parts can be made from 41XX mat’l and not hardened. That is certainly true for RC23.

The square is just under the 9mm flats so it is possible that a team can modify the existing shafts to mate with a 11/32" broached hole.

These prints were not made to correspond to my organized Plans A, B, & C so they are kind of a mish-mash. Pick and choose as you see fit.

Please check dimensions for us. Check fits and tolernances too if you know what you are doing in this area. Report errors and clarifications needed.

Note that the dowel pins on the existing carrier are undersized 4mm pins. Brian Orr (my long suffering friend who did these prints for me – staying to 11pm and his last day of working for Delphi – special thanks to Brian whi is not even a FIRST fanantic.) sized the hole for such undersized dowel pins. If you are going to use standard 4mm dowels, you may have to open up the holes appropriately – press fits are tricky be careful.

56mm Carrier and Shaft prints.ZIP (94.7 KB)


56mm Carrier and Shaft prints.ZIP (94.7 KB)

2

Great work on this, Dr. Joe. I know that the biggest priority is to get the 56mm gearboxes fixed, because the vast majority of teams will probably be using them for their drivetrains, but are there any similar plans in the wings for the 36mm box? I’m sure that there are a lot of teams worried about that one, too.

3

I don’t understand something - is the problem in the banebot gearboxes or is the problem in the transmissions?

Thanks,
-Guy

4

The Banebots 56mm Orbital gearbox is a transmission.

5

There is one solution you missed…

Junk the whole things and get Andy Mark shifters!

:slight_smile:

In all seriousness, we tried to make these things work out, but we aren’t willing to take a chance on them.

6

I think it would also be a good idea to include the 42mm gearboxes. A lot of teams are using them with the FP motors and with less than 3 weeks left, redesigning a mechanism may be too much for some teams. It’s a shame seeing some interesting designs or mechanisms be scrapped just because of one lousy carrier plate failing.:frowning:

7

I would like to thank Dr. Joe and all the others working to fix this problem. My team is using the banebots gearboxes and we feel it is too late in our design process to completely switch them out. We here at team 1676 anxiously await any solution to the problem. We would volunteer to help but unfortunatly our machining capabilities are limited and we usually have to outsource any custom cut or welded parts.

Thank you all once again for all of you working to rectify this situation.

8

A note to all the folks concerned about the 42mm and 36mm gearbox:

I appreciate your concern but the cases have almost nothing in common with eachother.

Let me explain. The 56mm gearbox is PURCHASED by FIRST to put in the kit. While teams can use them for any purpose they see fit, the main purpose that FIRST decided to buy them is to drive robot chassis for teams with scarces resources. The reason I have been loosing sleep is that I believe that a drivable robot is so fundamental to a successful FIRST season that the mission of FIRST was at risk of being damaged if teams were going to see these failures in competition.

Now to the 36mm gearboxes. These are gearboxes that Banebots DONATED to the kit. More than that, they donated an extra motor and other goodies. Beyond that, they offer for sale other things that make building robots easier and more robust.

There is literally NOTHING in the FIRST KOP that cannot be damaged if used beyond their limitations. All KOP items, actually, all items have limitations.

In my work to solve the 56mm Gearbox Double D issue, I found that I could extrapolate some of my results to the 36mm gearbox that Banebots donated to the KOP as well as the 42mm gearbox that Banebots sells. I THOUGHT I was doing the FIRST community a service. I was offering advice as to what I thought a conservative use of these parts would entail.

While I am happy that I have perhaps prevented some teams from designing a mechanism that will break. I regret that many teams have inferred from this several false ideas:

  • The 36mm and 42mm gearboxes are defective* – THIS IS FALSE
  • The 36mm and 42mm gearboxes are not extremely useful for a FIRST robot – THIS IS FALSE
  • FIRST and/or Banebots should do something to remove this limitation – THIS IS FALSE
    Perhaps my greatest regret is that my conservative estimates have caused significant financial damage to Banebots**. While I have not spoken to Banebots about specific numbers, it is clear that teams are mis-interpreting my conservative estimates to mean that they should not use Banebots’ 36mm and 42mm gearboxes at all. Nothing could be farther from the truth.

While teams are free to do whatever they want, I really really hope that teams will think and act rationally. I believe that if they do, they will end up buying quite a few products from Banebots. I think this will be a good thing for a number of reasons not the least of which that FIRST benefits from companies like AndyMark, IFI, and Banebots being around and supporting FIRST. Sales are what keeps a business in business. Sales to FIRST teams make donating stuff (or selling it at a deep discount) to the FIRST KOP a lot easier to fit into a business plan.

For what it is worth.

Joe J.

*believe it or not, a team actually tried to return the KOP 36mm gearbox claiming it was defective – this is insanity, not to mention about the most ungracious request I can imagine. Think about it, a company donates something to the FIRST KOP and a team has the audacity to request a cash refund because the part could, in theory, be broken. My blood boils!!!

**I will state again for the record, other than the fact that Robotic Amusements, Inc. uses Banebots as a supplier of some of its motors, I have no financial tie to Banebots at all. To be specific, they don’t pay me, I get no commision, I don’t get a discount or kickback on other sales, I am not a paid consultant for them, etc. etc.

9

Dr. Joe…

Team 166 has been following these threads and support everything you and all the other teams offering suggestions, perform testing, etc.

However, I believe that it would be a much easier task for those of us planning to manufacture new parts if we had drawings to use.

Are there any plans from BaneBots to supply the drawing for the carrier plate for the 56mm gearbox?

Please forgive me if this question has been answered in a previous thread. My time is limited to search for such information.

C. B. Petrovic - Team 166

10

Lockheed Martin seems to be in the mood to detemine the carrier plate alloy. I found a sympathetic lab manager that has at his disposal all the NDE equipment we use making the space shuttle external tank. Are people still interested in this?

11

Joe, Many thanks to you and others for your tireless efforts to remedy this situation.

I agree, and I am sure you are right about the solution. It seems everyone involved was trying to do good. I hope that any losses (both tangible and intangible) are minimized for all concerned parties.

12

Good Idea. It was in the works but I had to sleep last night and I had my job to do this morning. I attached a zip file to the first posting above.

13

Thanks but I think the need for mat’l reverse engineering has past. BUT… …you’re on the list. We will keep you in our pocket for a later date when we might just pull you out and use you.

Thanks again.

Joe J.

14

Thanks again for your hard work this week, Joe. Cool heads will prevail and I hope we don’t scare away suppliers…the more the better.

Thanks also for the drawings. I sort-of guessed from measurement, knowing it was metric, but that will help a lot. I am going to try to cut some test carriers out of O1 or A1 tool steel (similar in hardness to the 42mm carrier and capable of being tempered). It will mean hunting for waterjet time, which is tough because there is a class that uses them a lot right about this time of January/February. But I’ll let everone know how it goes.

15

Good work on the drawings Joe. Is there any reason why a single plate couldn’t be used for both the 4:1 and 3:1 ratios? Just rotate one bolt pattern by 36 degrees relative to the other. That might make the manufacturing of replacement plates more streamlined.

16

got em!

thanks again, Dr. Joe…

C.B. Petrovic - Team 166

17

Joe,

The material stated on your drawing. Is that the material we should make it out of, or the material it was made out of?

18

I would like to add that we used smaller Banebots transmissions last year for Mabuchi and Fisher Price motors and found the quality and service excellent.

19

There are many mat’ls that will work. I recommend that you talk to your source and pick one that has a RC23 hardness out of the shoot (i.e. without heat treating). The mat’l on the print is what I would do if I wanted to end up with a RC40 or so part. But you can relatively easily machine RC23 parts (or I should say, some machine shops can). Talk to the source about what they can get easily. If the mat’l is RC23 or higher, it should be fine.

Joe J.

20

The only reason not to is that if you are making a bunch of them the price will be a bit higher. For ones and twos, the set up cost would swamp the extra machine time so it is likely that you could get all 10 holes for almost the same price as just 5.

You pays your money and you gets your choice…

Joe J.