pic: 25lb Drivetrain

[Knufire]

loading photo ...

[cad321]

Is there a reason you chose not to use bearing blocks on your drive wheels? Although I haven't built a WCD before, my understanding is that their benefits far outweigh the cons (the only con in this I could see is possibly weight as you're limiting it to 25lbs).

[Sperkowsky]

By removing the crossmember 2x1s in favor of the bent sheet metal you are saving a bit of weight. But it is not as strong and you are more limited in mounting points. The few pounds its saves is just not worth it imo.

[Gregor]

"Taking a year off"

[asid61]

Originally Posted by cad321

Is there a reason you chose not to use bearing blocks on your drive wheels? Although I haven't built a WCD before, my understanding is that their benefits far outweigh the cons (the only con in this I could see is possibly weight as you're limiting it to 25lbs).

The other cons are cost, size, and time. If I had the gearboxes (and sheet metal brake) I could make this very quickly. It's possible to re-use the bearing blocks for sure, but the "set and forget" method of direct C-C chain is great.
From a machining standpoint, depending on which bearing blocks one is using, this can also be easier to machine than a "regular" WCD with bearing blocks.

[bkahl]

Originally Posted by Sperkowsky

By removing the crossmember 2x1s in favor of the bent sheet metal you are saving a bit of weight. But it is not as strong and you are more limited in mounting points. The few pounds its saves is just not worth it imo.

A cross section of a C is almost negligibly weaker than a box. In this scenario, a single piece belly ban and front and back members is actually extremely rigid.

On top of that, a single-piece bumper could add even more strength.

By your logic, would not all drives with a similar single-piece belly pan and cross-members be weak, and limited in mounting? IIRC 971 uses a similar style design. What about the kit frame? It uses C-channel cross-members too.

EDIT:

Originally Posted by cad321

Is there a reason you chose not to use bearing blocks on your drive wheels? Although I haven't built a WCD before, my understanding is that their benefits far outweigh the cons (the only con in this I could see is possibly weight as you're limiting it to 25lbs).

Another reason for this design may be for maintenance.

With the one-piece belly pan and members design, you "cap" the end of your drive tubes, making it hard to change a chain in an (unlikely) emergency. Being able to pull bearings from the tube may allow you to slide a chain in without taking the belly pan off (even though it still sounds like a nightmare).

[Knufire]

Originally Posted by cad321

Is there a reason you chose not to use bearing blocks on your drive wheels? Although I haven't built a WCD before, my understanding is that their benefits far outweigh the cons (the only con in this I could see is possibly weight as you're limiting it to 25lbs).

Yup!

There's two main benefits to using sliding bearing blocks: to better support the bearings and to be able to tension the chain. I'm not a big fan of adding tensioners where they might be unnecessary, and had pretty good luck running exact C-C #25 chain drivetrains in 2015 and 2016.

So the question is then, is putting the bearing straight in the tube sufficient support? This is game dependent; large impact loads on your wheels could cause the bearing holes to start to turn into ovals. However, for most games which have no or relatively small field obstacles, bearings straight in the tube should be fine.

[Harrison.Smith]

Originally Posted by Knufire

Yup!

There's two main benefits to using sliding bearing blocks: to better support the bearings and to be able to tension the chain. I'm not a big fan of adding tensioners where they might be unnecessary, and had pretty good luck running exact C-C #25 chain drivetrains in 2015 and 2016.

So the question is then, is putting the bearing straight in the tube sufficient support? This is game dependent; large impact loads on your wheels could cause the bearing holes to start to turn into ovals. However, for most games which have no or relatively small field obstacles, bearings straight in the tube should be fine.

Past 2 years we have ran WCDs. 2015 we ran C-C #25 and we had no outer support with the bearing going straight in the tube. That worked just fine, although RR wasn't very taxing on drives. This year we had a hybrid C-C #25 in tube and C-C #35 on the outside of the tube. We had 3/16 stiffener plates welded to the tube that had couterbores for the bearing. While the plates were there to stiffen the tube since we had a #35 chain run on the outside of the tube resulting in a large slot on the end of the tube, we had no problems with the bearing holes turning into the ovals that likely would have happened if the plates were not present.

[bkahl]

Originally Posted by Knufire

So the question is then, is putting the bearing straight in the tube sufficient support? This is game dependent; large impact loads on your wheels could cause the bearing holes to start to turn into ovals. However, for most games which have no or relatively small field obstacles, bearings straight in the tube should be fine.

https://www.chiefdelphi.com/forums/s...d.php?t=146519

[Knufire]

Originally Posted by bkahl

https://www.chiefdelphi.com/forums/s...d.php?t=146519

Lol, that's exactly the picture I was thinking of. Glad to see that people reported in that thread that bearings in .125" tube survived Stronghold.

[Chris is me]

Putting bearings straight into 1/8" wall tubing is just fine, it's only 1/16" wall tubing you really have to worry about ovalizing with. You should be fine there.

I'd consider riveting the bearings in instead of screws, just because I hate 4/40 screws and breaking taps and stuff. I guess if you have to change out a bearing there will be some rivet scraps inside your tube and that's kind of nasty.

Originally Posted by bkahl

A cross section of a C is almost negligibly weaker than a box. In this scenario, a single piece belly ban and front and back members is actually extremely rigid.

On top of that, a single-piece bumper could add even more strength.

By your logic, would not all drives with a similar single-piece belly pan and cross-members be weak, and limited in mounting? IIRC 971 uses a similar style design. What about the kit frame? It uses C-channel cross-members too.

I agree that the one piece belly pan / C-channel member will make this chassis likely rigid enough to perform well, particularly with stuff on top of it, but I wouldn't say a C-channel is "negligibly weaker" than tubing. It's certainly got substantially less resistance to bending. In this case, it's somewhat offset by the fact that it's got a giant bottom flange connecting it to the other side, and that the C-channel hugs the inside of the drive tube. This style of drive has certainly been battle-tested before.

I think this drive would get some strength benefit from making the top flange longer. Just an inch long is a little wimpy as a C-channel, and it could kink in the middle if hit really hard (and the bumpers aren't backed up with their own frame etc). I would just extend the whole flange to the length that is present at the points where it attaches to the frame. You're already using that material anyway; what's another few ounces per side to resist bending on impact a bit better. Especially if the belly pan is .090.

[Ty Tremblay]

I wouldn't even bother with holding the bearings in the tube. If you used screws and washers on the ends of the axles (or properly toleranced snap ring grooves) the bearings will have no where to go any way. 319 tapped the ends of our thunderhex shafts 1/4-20 and ran them all season without issues.

[JesseK]

The bellypan design is quite interesting. If it were the basis for a kitbot, it would turn the kitbot frame into a 7-piece rivet operation.
It's like the AMU chassis found a way to become WCD.

[Akash Rastogi]

Originally Posted by Sperkowsky

By removing the crossmember 2x1s in favor of the bent sheet metal you are saving a bit of weight. But it is not as strong and you are more limited in mounting points. The few pounds its saves is just not worth it imo.

Meh, I've done it before. You don't miss out on much if the rest of the bot is planned out with enough attachment points. And the rigidity is there with the bellypan.

[GeeTwo]

I do like the concept, and the weight economy! It's designs like this that make me want the team to invest in a sheet metal brake, probably even ahead of a lathe. If we just had the space for some real shop machines (32'x24' for storage and workshop and programming and maybe everything else).

While c-channel is not as strong as tube (especially in torsion, and significantly in around the axis perpendicular to the "missing" face), I do not see this as a showstopper - as previously noted, the 2014-2016 kit bots used folded c-channel to rave reviews. (3946 purchased some 2015 frames on closeout that we plan to use in 2017, or in 2018 on the low-percentage chance that we decide to go another route.)

There is plenty of room for attachment points. The top of the front and rear rails, and the portions of the tube which do not have chain behind them can be perforated as needed.

I understand that chain-in-tube is usually "set it and forget it" over the course of an FRC season, but with the capped tube ends, this design does not seem to have a good way to swap out the chains if disaster should strike. As a first mitigation, I would probably NOT perforate the drive tube for mount points, but rather mount a c-channel or thick-wall tube of versa-frame in a location which would allow a side tube, wheels, chains, gearbox, and motors to be removed as a single unit without removing anything else but the bumpers, or design "everything but" the drive chassis and things mounted to the belly pan to be easily removable to allow drive rail removal and repair.

Finally, This was not a design point, but someone brought it up: I don't see how this could work as a kit frame. Kits usually ship in a box roughly 36" x 8" x 8", but this belly pan unit would have to ship as something rather larger. Even more critically, there is no good way to make this so that a team can select "long/square/wide" using a band saw. (And OBTW, the kitbots since at least 2012 have a LOT more options than are listed in the manual, if you are willing to put in a bit of skull work.) If you skipped the belly pan, much more is possible, but that misses the key feature that makes this such a weight saver.