Understanding WCP Drive Train Calculator

[rlance]

http://i.imgur.com/5OnDLk1.png

I'm working with my team on gearing and determining the best ratio based on what they are wanting to achieve. I have a few questions about the results on the image above.

1. Motor amps @ Max - If that is above 40 amps do we just need to make sure we aren't giving full power to the motor at the wrong time (during initial acceleration)?

2. Top Speed - If we are building a robot for speed, what value should we be aiming for? I don't want to aim for 12 if that is not really that fast compared to other robots.

3. Pushing Force - If we are building for pushing power, what is a good number to shoot for?

4. Kinetic CoF - How is this number arrived at? We are plugging in the Static CoF using values we find for the wheel on Vex's site, but not sure about Kinetic.

Thanks in advance for the help.

[Knufire]

1. You're fine during acceleration; the motor is only stalled for an instant. You can see here from the 40A breaker datasheet that you can draw large amounts of currents for a short amount of time. You really just want to avoid any scenarios where you would be stalled for an extended period of time (a head-to-head pushing match where neither robot breaks traction, stalled against a wall, etc).

2. Max speed is a bit of a misnomer. Opinion on the ideal top speed varies greatly with team and with region. It is obviously beneficial to go faster, but there are drawbacks. The higher your top speed is, the slower you will accelerate. In addition, you'll draw more current, meaning you need to pay more attention to your wiring, battery charging, and battery health in order to minimize resistance before the motor.

The best advice I can give you is this; there are various robot acceleration models floating around on Chief Delphi. Based on the game and your strategy, pick a distance that you think would be the most traveled (e.g. from the edge of the defenses to the low goal) and pick your gear ratio to minimize the time to go this distance.

[InFlight]

I'm not a big fan of these calculators as you miss out on understanding the fundamentals equations of why you would want a particular ratio. These equations ignore any inefficiencies in the drive train, but give you a better insight into the tradeoffs.

Gear Box Selection

Calculating Wheel Speed and Gear Ratios

The distance covered by one rotation of a wheel is it circumference in feet.

Circumference = Pi * (wheel size in Inches)/(12 inch/foot)

Assuming the use of the normal CIM motors, 4455rpm is a typical design speed. The wheel speed in Feet per Second becomes:

Speed= (4455 rpm x Wheel Size x 3.14)/(60sec/min x Gear_Ratio x 12in/foot)

To identify the necessary gear ratio to achieve a desired speed, we can rearrange terms for-

Ratio= (4455 rpm x Wheel Size x 3.14)/(60sec/min x Speed x 12in/foot)

As an example, assume the use of a two speed gearbox with 8 inch tires.
We would like a high speed of around 16 feet/sec to move very quickly.

High Gear = (4455 rpm x 8" x 3.14)/(60sec/min x 16 ft/sec x 12in/foot)=9.7

We would also like a low speed of 4 feet per second to have a good pushing gear.

Low Gear = (4455 rpm x 8" x 3.14)/(60sec/min x 4 ft/sec x 12in/foot)=38.8

Limiting ourselves to commonly available to two speed gearboxes , the closest one to our desired specs is Vex 3 CIM Ball Shifter with the 20/64 output gearing that provides 9.07 & 33.33 output ratios. This gearbox option would provide:

High Speed= (4455 rpm x 8" x 3.14)/(60sec/min x 9.07 x 12in/foot) = 17.1 feet/sec

Low Speed= (4455 rpm x 8" x 3.14)/(60sec/min x 33.3 x 12in/foot) = 4.7 feet/sec

If we were selecting a single speed gear box for 8 inch tires

Assume a 14 ft/sec speed; which is a ratio of 11.1 using the methods outlined above.

There are three different single speed AndyMark Gearboxes that have 10.71 output ratio options, which would result in a predicted speed of 14.5 feet per second.

Torque and Acceleration are proportional to the Gear Ratio. Speed is inversely proportional to the gear ratio. In an open field you may favor speed. In a congested field with shorter travel distances, acceleration is more important. So choose your top speed for the gearboxes accordingly.

[Andrew_L]

I'm not a big fan of these calculators as you miss out on understanding the fundamentals equations of why you would want a particular ratio. These equations ignore any inefficiencies in the drive train, but give you a better insight into the tradeoffs.

It's definitely good to understand the math behind the calculators, but I don't think you're giving the calculators enough credit. Drivetrain calculators are an effective way of quickly seeing the visualization of a drivetrain as you change specific variables, something that would take much longer when doing the math by hand.

I've been designing drivetrains for 6 years now and while I *can* do all of the equations, I always have a tab open of JVN's Mechanical Design Calculator and the WestCoast Products Drivetrain calculator. These calculators are so useful that learning how to utilize them is an invaluable tool for any designer.

[GeeTwo]

I'm not a big fan of these calculators as you miss out on understanding the fundamentals equations of why you would want a particular ratio. These equations ignore any inefficiencies in the drive train, but give you a better insight into the tradeoffs.

Being a big fan of "use the source", I have learned a lot about what is important in the FRC drive trains and other systems by ::rtm:: reading and interpreting the formulas in various calculators.:] It's very convenient to be able to do the calculations in your head (to around two decimal places) to be able to figure out what is and is not feasible on the spot, but before any money is spent, yes, I go back to a spreadsheet (sometimes my own, sometimes a published one) where I can examine all the steps.

[Oblarg]

One major flaw of the WCP drive calculator that you should take note of: the "max pushing force" and "max current draw" calculations both use static wheel COF instead of dynamic. This is very misleading, as if you're actually traction-limited, you'll be spinning the wheels and the static COF only matters for a moment until the wheels slip. So, if you're trying to see if your current draw in a pushing match is sustainable without tripping a breaker, the numbers they give are not correct.

I emailed them about this some years back, but it doesn't seem that they ever got around to fixing it.

OP: As for your questions, there is no one-size-fits-all answer to any of them. You have to wrap your head around the concepts behind that calculator - the tool is there to save you the effort of doing the calculations by hand, but if you don't have at least a heuristic understanding of what the calculator is doing then it's very hard to use it properly.

That said, here's my best attempt at giving a short-ish answer to each of your questions. Please, do not just read these and ignore what I said above. They are not absolute guidelines, and are meant as prompts for further thinking.

1) It depends. There are several concerns due to excessive power draw; the most pressing are browning out the roborio or tripping the main breaker, but past that you have the individual 40-amp drive breakers, and also overall battery usage (if you have a power-hungry robot this can indeed be a concern). Each of these requires separate consideration. If you are only running a 4-CIM drive, you will likely only have to worry about the drive breakers; if you are running a 6-CIM drive, the others come into play.

2) This depends on the game. A higher top speed means slower acceleration. In a field like 2016, where there was not much open running room for a robot, gearing a single-speed robot for faster than 12 fps or so (YMMV pending drive efficiency and other factors) would have likely been a waste as you'd rarely actually reach that speed (shifters can change that, though it either requires auto-shifting code or very skilled drivers to actually take advantage of them for acceleration). In a game like 2014, with a wide open field, a high top speed could be very useful.

3) "Pushing force" isn't really a value people tend to explicitly design FRC drives around - you rarely "shoot for" a specific value here. Again, as well, this depends on the context - if you're a defensive bot in a contact-heavy game (2014?), being able to push can be extremely valuable. In other contexts, it's completely unimportant. If you do decide that you need to optimize pushing ability (I've never actually done this, but I can imagine some contexts in which you might want to), the "pushing force" value on the calculator isn't really that important - you make your robot as heavy as possible, pick the traction-y-est wheels possible (within reason - I'm sure you could get fabulous pushing performance from the soft rubber BaneBots wheels, but I doubt the field crew would take kindly to the resulting rubber shavings left on the field and you'd probably have to swap every match) and then pick a gearing that will let you push constantly under those parameters without causing any power issues. This will almost certainly require a 2-speed gearbox, because the gearing that allows you to do that is likely to not be very mobile.

4) This, of course, depends on which wheel you're using - if you dig around CD with the search function you can likely find various test results for various wheels that members have done, though I seem to recall being surprised at how much the quoted values from different sources vary when I've looked. I will say that, based on toying with the calculator and comparing it to my actual experiences, .7 is a pretty reasonable guess for AndyMark HiGrip wheels and similar.

If you'd like, you can send me a PM and we can discuss more in-depth.

[Chris is me]

One major flaw of the WCP drive calculator that you should take note of: the "max pushing force" and "max current draw" calculations both use static wheel COF instead of dynamic. This is very misleading, as if you're actually traction-limited, you'll be spinning the wheels and the static COF only matters for a moment until the wheels slip. So, if you're trying to see if your current draw in a pushing match is sustainable without tripping a breaker, the numbers they give are not correct.

This isn't quite as black and white as you make it out to be.

Three cases to consider:
1. Less than your maximum pushing force is required to displace whatever you are pushing. In this case, your wheels don't slip, they start moving before that happens.
2. More than your maximum pushing force is required. You're already not pushing them, whatever. In fact once you start to slip you'll draw slightly less current than the max!
3. Right on the borderline. Behavior varies around here; the winner of the pushing match if head to head is usually who carried more momentum into the push.

The bigger deal that isn't modeled is the effects of voltage drop under load on pushing force and current draw. This is significant.

I think it's plenty reasonable to model at static CoF - unless you have a low gear that's mega low and you're pushing boulders (grippy boulders, made of rock, not foam), your wheels probably won't slip before you move anyway.

[Oblarg]

This isn't quite as black and white as you make it out to be.

Three cases to consider:
1. Less than your maximum pushing force is required to displace whatever you are pushing. In this case, your wheels don't slip, they start moving before that happens.
2. More than your maximum pushing force is required. You're already not pushing them, whatever. In fact once you start to slip you'll draw slightly less current than the max!
3. Right on the borderline. Behavior varies around here; the winner of the pushing match if head to head is usually who carried more momentum into the push.

The bigger deal that isn't modeled is the effects of voltage drop under load on pushing force and current draw. This is significant.

I think it's plenty reasonable to model at static CoF - unless you have a low gear that's mega low and you're pushing boulders (grippy boulders, made of rock, not foam), your wheels probably won't slip before you move anyway.

The thing is, in my experience, your 'case 2' is by far the most common.

Even if your robot can move, you can (and will) still slip the wheels if your motor torque is high enough to break static friction, unless you are being very careful with the throttle. In fact, there is actually little/no difference in the first few moments between accelerating full-throttle from a dead stop, and pushing against an immovable wall. In both cases, your wheels will slip if your robot is traction-limited.

You are correct that when the wheels slip, you draw less current than the max - and it's not "slightly less," either. The difference is significant. This is why it's so dangerous to be right near the traction-limited threshold - if you are just over it and you push, your wheels will slip and you will likely be fine; but if you're just under it, your motors will stall and you will not. There is a sharp discontinuity in the current draw you see in a pushing match between the two, and it corresponds precisely to the difference between static and dynamic wheel COF.

4464 discovered this firsthand in 2014 between our two regionals - we added some just enough weight between them to cause us to be motor-limited when our battery wasn't quite full, and the result was slew of main breaker trips in the second when in the first we had had almost no problems.

[Chris is me]

The thing is, in my experience, your 'case 2' is by far the most common.

Even if your robot can move, you can (and will) still slip the wheels if your motor torque is high enough to break static friction, unless you are being very careful with the throttle. .

If it's case 2, it doesn't matter, you're not pushing them even if your wheels didn't slip.

If your robot can move the load reasonably well (and low gear isn't unreasonably low), then no, the wheels aren't slipping - the load is moving instead. Except for the edge case where you are just barely able to move the load you are trying to push, the wheels apply force, and the robot+load start to move before the wheel slips. The motor isn't stalling anymore, it's moving, the torque is being "used up" to make the robot move.

[Oblarg]

If it's case 2, it doesn't matter, you're not pushing them even if your wheels didn't slip.

If your robot can move the load reasonably well (and low gear isn't unreasonably low), then no, the wheels aren't slipping - the load is moving instead. Except for the edge case where you are just barely able to move the load you are trying to push, the wheels apply force, and the robot+load start to move before the wheel slips. The motor isn't stalling anymore, it's moving, the torque is being "used up" to make the robot move.

That's not how it works. If the torque output is high enough to break static friction, the robot moves *and* the wheels slip. Your effective acceleration will be limited by the dynamic COF of the wheel. This is the same reason we have antilock braking systems in cars.

You can get a robot and test this, if you like.

[Chris is me]

That's not how it works. If the torque output is high enough to break static friction, the robot moves *and* the wheels slip.

You can get a robot and test this, if you like.

It depends on what current load is required to break traction. Almost every robot can break traction at some current below stall, but wheel slip is negligible when accelerating from a full stop across a range of gear ratios. (It's probably instantaneously slipping, but not really for long enough to matter) If you're traction limited at, e.g., 60% of stall or ~70 amps per CIM (lots of faster single speeds), you'll really never slip in a pushing match that you're winning. If you're traction limited at 15% of stall or ~20 amps per CIM, you're probably burning out every time you gun the throttle. Most low gears in FRC seem to aim for traction limited at anywhere from 40 to 55 amps per CIM, and I just haven't observed slipping in pushing matches the robot handily wins. The load on the wheels is relieved to some extent when the load starts moving / slipping, and momentum helps too.

228's robot this year was 4 cim / 2 mini, geared to like 7.5 FPS low gear, and tread-cut Colson wheels. As long as the drive didn't ride up on the pushing target (grumble grumble), it would push things like AM wheel robots without the Colsons slipping noticeably. And that's traction limited at like, 35 amps per CIM? maybe 40? I can't remember off the top of my head, but less than I expected.

[Oblarg]

If you're traction limited at, e.g., 60% of stall or ~70 amps per CIM (lots of faster single speeds), you'll really never slip in a pushing match that you're winning.

This isn't what I've experienced. Now, I will admit that my teams rarely use super-high-traction wheels (AndyMark HiGrip and unmodded Colsons, generally), so that could be a contributing factor.

Mathematically, at full throttle you'll slip the wheels at any speed below where the motor's torque value exceeds the torque generated by the static friction of the wheel.

If you're traction-limited at 60% of stall, this means any speed below 40% of max (assuming a linear torque curve). This is probably not quite right, because of voltage drop and other issues, but I see a *lot* of pushing matches where the movement in either direction is clearly a lot slower than that.