Team 4592 Offseason Drivetrain

[thatprogrammer]

I've worked on designing this drive train for the last few days as a potential chassis we may be using next year. I've only used cad for a week, so there may some errors (I do not believe the top holes on the bellypan and the holes of the drive modules are connected for example).
Features:
C-C chain in tube design
Removable Modules for easy replacement of a damaged gearbox or drive shaft
Ability to use same module on both sides (no left or right hand versions)

Please look over the drive train and provide feedback!

[Chris Endres]

The overall design looks good, we built a similar design a few years back for a fall project and it worked well.

My first question: Why chain? I don't want to start any fights, but I personally feel that belts and pulleys would better in this situation. This is because getting the chain tension properly may be a problem, and making a dead axle hole with a CNC router or water jet or whatever you are using to cut your sheet metal is pretty easy and precise.

Primary concern: The side plates cannot be made the way they are now because of the bending geometry. You should redesign so there are two plates on each side that hold the wheel (basically split the middle plate in two). Of course, this would kind of eliminate the "removable module" idea, but this current part is not bendable, to my knowledge.

Another concern: You may want to look into making the front/back/belly pan part into three parts, you should check your bending capabilities for tolerances and actual production of the part (some brakes may not be able to bend the belly pan that way).

EDIT: You can still have "no right or left hand versions" for the side plates if designed correctly.

[thatprogrammer]

Originally Posted by Chris Endres

The overall design looks good, we built a similar design a few years back for a fall project and it worked well.

My first question: Why chain? I don't want to start any fights, but I personally feel that belts and pulleys would better in this situation. This is because getting the chain tension properly may be a problem, and making a dead axle hole with a CNC router or water jet or whatever you are using to cut your sheet metal is pretty easy and precise.

Primary concern: The side plates cannot be made the way they are now because of the bending geometry. You should redesign so there are two plates on each side that hold the wheel (basically split the middle plate in two). Of course, this would kind of eliminate the "removable module" idea, but this current part is not bendable, to my knowledge.

Another concern: You may want to look into making the front/back/belly pan part into three parts, you should check your bending capabilities for tolerances and actual production of the part (some brakes may not be able to bend the belly pan that way).

EDIT: You can still have "no right or left hand versions" for the side plates if designed correctly.

Thanks for the feedback! I used chain over belt because the thickness of pulleys would make the modules very thick, limiting bellypan size. I'm attempting to get a c-c chain run with this, preventing the need for a tensioner (I know 846 has run this in the past).
Are you sure the module appears impossible to bend? I designed it such that our sponsor could do the two center bends and then the edges. If the current configuration is impossible, I'll work on changing it!
Would you like CAD files to further examine the drivetrain?

[Chris Endres]

Originally Posted by thatprogrammer

Thanks for the feedback! I used chain over belt because the thickness of pulleys would make the modules very thick, limiting bellypan size. I'm attempting to get a c-c chain run with this, preventing the need for a tensioner (I know 846 has run this in the past).
Are you sure the module appears impossible to bend? I designed it such that our sponsor could do the two center bends and then the edges. If the current configuration is impossible, I'll work on changing it!
Would you like CAD files to further examine the drivetrain?

I have never seen a part like that be made, I would be really surprised to see a manufacturer can bend a part like that.

In terms of the chain, you may want to design this for the compatibility of both, it's a good designing standard for first iterations to have a backup plan, it would suck to make this and realize it doesn't work like planned.

Also, you can use this calculator to get close to a perfect tension, I know 118 used this method last year.
http://www.botlanta.org/converters/d.../sprocket.html

[thinker&planner]

Unless your sponsor has a break that is less than ~1" tall (min distance between bearing plates) and can extend 4" (height from bellypan to top of tube) into a piece, they cannot bend that piece.

If you make the side "tubes" that the bearings sit in two separate pieces and overlapped them at the top and bottom, it would work. See Robowranglers' X009 Design Log.

Otherwise, looking pretty good. Just make sure that you leave space to get the sprockets and chain inside that side rail for installation and maintenance (I would make a hole in the belly pan in front of and behind the tube).

[asid61]

Originally Posted by Chris Endres

I have never seen a part like that be made, I would be really surprised to see a manufacturer can bend a part like that.

In terms of the chain, you may want to design this for the compatibility of both, it's a good designing standard for first iterations to have a backup plan, it would suck to make this and realize it doesn't work like planned.

Also, you can use this calculator to get close to a perfect tension, I know 118 used this method last year.
http://www.botlanta.org/converters/d.../sprocket.html

When going direct C-C on chain, make sure that your sheet metal sponsor can hold those tolerances.
Also for FRC, you should add around 0.017 - 0.019" to your C-C distance due to manufacturing tolerances on the sprockets and chain. See this thread: http://www.chiefdelphi.com/forums/sh...4&postcount=26

[Electronica1]

The important trick for sheet metal is understanding what your manufacturer can do. Talk to them and figure out what they can and can not do. It wouldn't be a bad idea to show them this cad and ask for feedback on it. (although, they will probably ask you break that channel into two separate parts)

[JVN]

Looks good!

Are those side "pontoons" each one large piece? If so, the sheet-metal bending may be touch to accomplish. It is tricky to get a die into such a tall "U" section. If these do get bent, you may find it is a struggle keep bearing holes aligned axially between the inside/outside of these U's.

The belly pan lightening pattern isn't very efficient. Is this being laser cut? Turret punched? You can potentially save a lot more weight out of it -- either by moving to a more "traditional" truss pattern, or just adding additional holes and/or smaller holes within your current pattern.

I'm always a big fan of the front/back angles.

How tall is the frame? Depending on wheel/gearbox config, it looks like you could tighten it up a bit.

Good start -- keep iterating! Each one will get better.

[philso]

Originally Posted by JVN

Looks good!

Are those side "pontoons" each one large piece? If so, the sheet-metal bending may be touch to accomplish. It is tricky to get a die into such a tall "U" section. If these do get bent, you may find it is a struggle keep bearing holes aligned axially between the inside/outside of these U's.

The belly pan lightening pattern isn't very efficient. Is this being laser cut? Turret punched? You can potentially save a lot more weight out of it -- either by moving to a more "traditional" truss pattern, or just adding additional holes and/or smaller holes within your current pattern.

I'm always a big fan of the front/back angles.

How tall is the frame? Depending on wheel/gearbox config, it looks like you could tighten it up a bit.

Good start -- keep iterating! Each one will get better.

This is my concern also. How tall are the pontoons on each side? could you use 2" x 1" or 3" x 1" tubing instead of sheet metal. The bearing holes on both sides could be cut in one operation and it should be relatively easy to get them to line up. Depending on the wall thickness of the tubing you choose, you would end up with a similar weight to your current design.

The short vertical surface at each end of the belly pan may be difficult to make, depending on the sheet metal brake available to you. The design with both ends bent up means that the belly pan and the pontoons must be made to match very closely or the parts will not fit. It may be better to make the two ends as separate (identical) pieces as Chris has already suggested. The end pieces can be riveted onto the flat bottom piece. You may want to manufacture the parts so that there are only holes on the ends or on the flat middle part then match drill and rivet so that it all self-adjusts for the real-world manufacturing tolerances.

As most of the other people answering you have said, you need to get know what your sheet metal supplier is capable of and design around that. The Mechanical Designers where I work do this as a matter of course.

[Munchskull]

Originally Posted by asid61

When going direct C-C on chain, make sure that your sheet metal sponsor can hold those tolerances.
Also for FRC, you should add around 0.017 - 0.019" to your C-C distance due to manufacturing tolerances on the sprockets and chain. See this thread: http://www.chiefdelphi.com/forums/sh...4&postcount=26

Is that added distance for #35 or #25 chain?

[notmattlythgoe]

Originally Posted by Munchskull

Is that added distance for #35 or #25 chain?

#25 we have not figured it out for #35 yet.

[Munchskull]

Originally Posted by notmattlythgoe

#25 we have not figured it out for #35 yet.

Any resources out there of how to calculate it?

[notmattlythgoe]

Originally Posted by Munchskull

Any resources out there of how to calculate it?

We did it by testing it on actual #25 chain. The same process from the linked thread could be used on a #35 chain setup.

[Knufire]

Originally Posted by Munchskull

Any resources out there of how to calculate it?

Paul Copioli has a chain calculator out there that had a 0.018" addition for #25 and 0.012" for #35.

http://www.chiefdelphi.com/media/papers/3166