FRC Blogged-Kit of Parts Drive System Option

[coalhot]

Interesting, I wonder if this is any indication on next years game.

Looks interesting though...

[Akash Rastogi]

Excited to see what the new kit frame looks like. Making superstructure attachment easier might make more teams want to stick to the kitbot.

Nice work AM!

[AllenGregoryIV]

I'm guessing the standard 12.75:1 toughbox mini with 6" wheels. That's 10.9fps unloaded and 8.83fps with standard JVN loads.

I really do wish they would give a bit more information. I always like having a KOP drive to build so my team can be trained to help rookies. If I know it was simple enough to not need it I would rather have the voucher.

Giving a direct drive system to teams is very interesting.

[D.Allred]

Originally Posted by AllenGregoryIV

I'm guessing the standard 12.75:1 toughbox mini with 6" wheels. That's 10.9fps unloaded and 8.83fps with standard JVN loads.

10.7:1 looks like the correct option.

Anyway, I'm interested to see a sheet metal chassis that is long or wide out of the box.

[Ether]

Originally Posted by AllenGregoryIV

I'm guessing the standard 12.75:1 toughbox mini with 6" wheels. That's 10.9fps unloaded and 8.83fps with standard JVN loads.

For students with inquiring minds...

In the spreadsheet, the 8.83 number is calculated directly from the 10.9 number by multiplying by a "Speed Loss Constant" (default is 81%).

This constant is not computed within the spreadsheet from user-entered info about physical loads and motive forces in the drivetrain: you get the same 8.83 answer from the spreadsheet if you change the number of motors, stall torque of motors, the mass of the robot, or gearbox/drivetrain friction.

The best way to determine what value you should use for the "Speed Loss Constant" for your drivetrain is to build the drivetrain and measure its top speed, then divide that by the theoretical unloaded speed.

[Whippet]

Perhaps they wanted to make it easier to attach superstructures because this year's superstructures will need a lot of structural support to do some heavy lifting? I'm calling stacking game.

[Ether]

Originally Posted by Ether

This constant is not computed within the spreadsheet from user-entered info about physical loads and motive forces in the drivetrain: you get the same 8.83 answer from the spreadsheet if you change the number of motors, stall torque of motors, the mass of the robot, or gearbox/drivetrain friction

Also worth noting: in the real world, the % of "speed loss" is not constant when changing gear ratios either:

Team 1678 did some testing earlier this year on a bot with a 2-speed gearbox. The "Speed Loss Constant" was 92% in low gear and 72% in high gear.

[T^2]

Originally Posted by Ether

Also worth noting: in the real world, the % of "speed loss" is not constant when changing gear ratios either:

Team 1678 did some testing earlier this year on a bot with a 2-speed gearbox. The "Speed Loss Constant" was 92% in low gear and 72% in high gear.

Ether is correct on most accounts. Keep in mind that JVN's "speed loss constant" comes after a "drivetrain efficiency" (i.e. losses from gearing), therefore the 81% number cannot be attributed to gearbox losses. The testing we did earlier this year found the overall speed loss from the theoretical free speed of the motors. The numbers we found make intuitive sense, given that motor resistance, friction, rolling resistance, and air drag all tend to increase as speed increases.

My conclusion now is the same as it was then: test your drivetrain instead of blowing smoke from your behind about the theoretical speed you think you'll get.

[Ether]

Originally Posted by T^2

Keep in mind that JVN's "speed loss constant" comes after a "drivetrain efficiency" (i.e. losses from gearing), therefore the 81% number cannot be attributed to gearbox losses.

The Speed Loss Constant does not come "after" the drivetrain efficiency (at least not in the 2013 version). It is completely independent. See below:

The testing we did earlier this year found the overall speed loss from the theoretical free speed of the motors.

^^This is exactly what the "Speed Loss Constant" in the JVN spreadsheet is.

My conclusion now is the same as it was then: test your drivetrain instead of blowing smoke from your behind about the theoretical speed you think you'll get.

I love test data too.

But I still think it may be possible to develop a physics-based model that will reasonably predict overall drivetrain performance -- including not only top speed but also time to reach a desired speed, time to reach a desired distance, and accel, speed, distance, motor amps, motor volts, and Coulomb consumption versus time -- based on a priori estimates of a limited number of parameters .