pic: Offseason 8W Tank Drive

[Jared]

Originally Posted by Oblarg

What program are you using for those drive numbers? AFAIK, it's pretty much impossible to pull 90 amps per CIM on a 6CIM drive under any circumstance with the batteries we use in FRC.

He's using JVN's design calculator. It just calculates out how much torque you'd need to slip the wheels with the gear reduction, then, from the motor curve, it figures out the current draw per motor. For high gear, voltage drop will play a huge role, and I don't think he'd be able to get the wheels to slip at all. The current/motor number is more important in low gear.

FWIW, we found that the 1.3 CoF was a little on the high side for normal traction wheels.

[Oblarg]

Originally Posted by Jared

He's using JVN's design calculator. It just calculates out how much torque you'd need to slip the wheels with the gear reduction, then, from the motor curve, it figures out the current draw per motor. For high gear, voltage drop will play a huge role, and I don't think he'd be able to get the wheels to slip at all. The current/motor number is more important in low gear.

FWIW, we found that the 1.3 CoF was a little on the high side for normal traction wheels.

Thanks. Not taking battery voltage drop into account would explain that pretty well.

I know the calculator on the WCP website does take battery voltage drop into account, which is nice, but their "max pushing force" calculation uses the static COF even for a traction-limited drive, which is not-so-nice.

[Michael Hill]

Originally Posted by Oblarg

Thanks. Not taking battery voltage drop into account would explain that pretty well.

I know the calculator on the WCP website does take battery voltage drop into account, which is nice, but their "max pushing force" calculation uses the static COF even for a traction-limited drive, which is not-so-nice.

Try my calculator out.

http://www.chiefdelphi.com/media/papers/3038

[Knufire]

A quick question, the sheet piece as shown is made out of 0.090" Al 5052-H32. I know other teams, such as 33 and 67, have had success with thinner sheets. Would trimming down to 0.063" be a wise decision?

[Michael Hill]

Originally Posted by Knufire

A quick question, the sheet piece as shown is made out of 0.090" Al 5052-H32. I know other teams, such as 33 and 67, have had success with thinner sheets. Would trimming down to 0.063" be a wise decision?

You can go even thinner. We're experimenting with 0.050" 2024-T3 this offseason, and it should actually be stronger than .090" 5052-H32. We get this added strength by not drilling lightening holes. Your robot is really only as strong as the smallest cross-sectional area (in tension and compression). So by eliminating lightening holes, the smallest cross-sectional area ends up where the axle holes are. It ends up almost the same weight, but almost twice as strong (ballparking, of course).

[asid61]

Originally Posted by Michael Hill

You can go even thinner. We're experimenting with 0.050" 2024-T3 this offseason, and it should actually be stronger than .090" 5052-H32. We get this added strength by not drilling lightening holes. Your robot is really only as strong as the smallest cross-sectional area (in tension and compression). So by eliminating lightening holes, the smallest cross-sectional area ends up where the axle holes are. It ends up almost the same weight, but almost twice as strong (ballparking, of course).

Is 2024 easy to bend?

[Michael Hill]

Originally Posted by asid61

Is 2024 easy to bend?

We use a 0.125" bend radius. It's a bit tighter than aviation requires, but it's good enough for FRC, especially if you bend against the grain. We've reached the limit of our brake, but fortunately it's big enough for a robot.

[asid61]

Originally Posted by Michael Hill

We use a 0.125" bend radius. It's a bit tighter than aviation requires, but it's good enough for FRC, especially if you bend against the grain. We've reached the limit of our brake, but fortunately it's big enough for a robot.

Oh, very good then. I was worried the stiffness would make the bend radius unusable. I'll have to ask our sheet metal guy if he can do 2024.

[Knufire]

Here's the current revision of the drive. The only main differences are the lightened bellypan and less aggressive pocketing on the front/back rail. Any last tips? I've debated elongating the top end rail flanges and possibly putting another flange down the back as 971 does.

Another change (not shown) is changing the outer gearbox bearings from round to hex. This allows the intermediate shafts to be pulled out without having to disassemble the outer drive rail.

[JesseK]

Hmm, 971's bellypan extends to the outer rails. They also have plates on the top of the rails for added strength. Have you considered doing this with your design, at least in the middle of the rails?

I'm no ME, but the proximity of the 4 rivets without any other support raises an eyebrow. The likelyhood of the sheet metal shearing or holes elongating (making the drive not square) seems pretty high.

[Knufire]

Originally Posted by JesseK

Hmm, 971's bellypan extends to the outer rails. They also have plates on the top of the rails for added strength. Have you considered doing this with your design, at least in the middle of the rails?

I'm no ME, but the proximity of the 4 rivets without any other support raises an eyebrow. The likelyhood of the sheet metal shearing or holes elongating (making the drive not square) seems pretty high.

I had a version where the bellypan extended to the frame perimeter. Since the wheels pairs are so close together and the chain hangs below the button of the frame, by the time you have wheel and chain clearance cutouts, you only have the middle section underneath the gearbox left, which makes it significantly harder to access the gearbox.



I've often done two rivets on a diagonal in a 1" square, but not 4. That I pretty much copied from the VexPro Drive in a Day, which I believe is bolted together. I'll definitely consider that.

The dead axles and gearbox shafts should help transfer the load from the outside drive rail to the inside drive rail on a hard side impact. Also, the plan for superstructure mounting is a 2"x1" box crossbar, which should help as well. I'm thinking about ways to mount the bumper that would increase the strength of the frame, as 33 does.

[Oblarg]

Originally Posted by Michael Hill

Try my calculator out.

http://www.chiefdelphi.com/media/papers/3038

This seems to be reasonably close to the experimental data I have, and all the behavior seems correct. Thanks for sharing, I'll be sure to keep this around.

Just one comment: Your "combined motor stall current" figure seems to be purely determined by motor selection, as it doesn't vary with changes to internal battery resistance or circuit resistance. The current displayed on the "stall conditions" graph, on the other hand, does seem to be calculated from the relevant values. Is there any reason for this?

[Michael Hill]

Originally Posted by Oblarg

This seems to be reasonably close to the experimental data I have, and all the behavior seems correct. Thanks for sharing, I'll be sure to keep this around.

Just one comment: Your "combined motor stall current" figure seems to be purely determined by motor selection, as it doesn't vary with changes to internal battery resistance or circuit resistance. The current displayed on the "stall conditions" graph, on the other hand, does seem to be calculated from the relevant values. Is there any reason for this?

That stall current is the "spec" current at 12V. This is just like the motor performance data we are given, but instead for a single motor, it is an equivalent for multiple motors. It is based on Paul Copioli's calculation in his "Useful Calculations" spreadsheet.

http://www.fightingpi.org/Resources/...culations.xlsm