The Ultimate FRC drive train

Technical Technical Discussion
1

what would the ultimate FRC drive train be? your wildest ideas are welcome.

edit: serious answers to

2 Likes
2
27 Likes
3

You might want to take a look at team 148’s x000 projects. It’s basically on offseason drivetrain design challenge. Their latest was x019, made in the 2019 offseason.

Also, 2910 has a really nice swerve drive setup.

But for the ultimate FRC drivetrain, it has to be the AM14u series. No other drivetrain has the history, versatility, or ease of setup all in one.

6 Likes
4

This will vary from year to year.

If engineering time and cost is no constraint and assuming a flat field. i’d say that a swerve drive is probably the “Ultimate FRC Drivetrain” style. The SDS MK4i or the WCP SwerveX modules connected with 2x1 versa-versa frame probably comes close to the ultimate drivetrain.

An argument could also be made for a lightweight west coast drive. You will be able to save a few pounds, and PDP slots with this style, leaving more possibilities for the stuff on top of the robot.

Now if you factor in engineering time and cost, the KOP chassis really hits a sweet spot. Especially with the upgrades available from Andymark. My favorite would be an 8 wheel setup. Performance wheels are more “ultimate”, but I hate wheel maintenance, so i’d probably just opt for the good 'ole white hi-grip wheels.

If you have the extra engineering/machining resources, a custom tank is probably a okay plan, if you know what you are doing, and have done something like it on the off season.

If you really have the extra engineering resources, and have done it off season, a COTS swerve has some nice advantages.

If you have so many engineering resources/experience that you don’t care what I think…maybe a custom swerve is for you :slight_smile:

Now for my fun one: I really think omni wheels on a swerve drive has some really interesting properties…

8 Likes
5

Continue…

9 Likes
6

I think decagonal wheels might be worth investigating.

16 Likes
7

Ahh…So I will preface this with…I have not done all the math…I could be making mistakes on some of the concepts. I’ll also say this concepts is probably very dumb, but interesting.

I initially thought of this as an “extension” to running an all-omni tank drive. With an all omni tank drive, an opposing defender T-boning you will not slow you down…only shift you to the side. I believe a few teams did this in 2017 with some success. You get a similar effect with the X019 that @Berkeli shared.

So…if you had omni’s on a swerve drive, you could control how much pressure you would put on an opposing defender as you pass. The opposing robot would not know what happens if they t-boned you. I believe you could even redirect their momentum hitting you to make yourself go faster (but this is where more math is required).

That was the initial thought, but I got to thinking about it further, and realized there was effectively a CVT built in to this system, and a theoretically infinite top speed (assuming 100% efficiency in the rollers).

Let me explain: If all 4 wheels face the direction you want to go, your top speed is calculated the same way it is on every robot. If you begin to rotate your wheels inward, you can increase that top speed. If you get to 45 degrees, you top speed is now sqrt(2)x(initial top speed). This happens with an H-drive going diagonally. Keep going, and your top speed can go to infinity. I’ve only convinced myself this works if you are going forward, backwards, left, or right. Not sure if this works in all directions.

So yeah…some “neat” properties.

11 Likes
8

It’s amazing to see how far FRC has advanced in just a decade! And to think these robots are designed by high school students… just wow :exploding_head:. If this is innovative design update is any indication of our future engineers, we’re all in good hands.

71 Likes
9

41 Likes
10

39 Likes
11

my teams robot (4265’s) may have skewed my perspective on the rarity of swerve.
links to our robot driving: Shifting Swerve? #Shorts - YouTube and It drives!!! - YouTube

12

have these results been shown in practice? I just find that really counter intuitive.

edit: some drawing would be appreciated.

2 Likes
13

In practice holonomic (corner omni) drivetrains are basically a fixed-wheel version of this. Their speed in the ordinal directions is sqrt(2)*wheel speed.

It’s really funny that omni swerve got brought up because I was just joking around about it with my team a few days ago. I’m sure it’s completely useless on an actual FRC robot, but you DO get CVT essentially for free.

14

so the x drive goes FASTER than a tank? if anything I thought they would be slower.

15

In order for it to actually achieve this the rollers have to roll rather than slide on the ground. YMMV in practice.

1 Like
16

This deck has some interesting drives in among them the grasshopper or butterfly. I bring them up because of the pneumatic actuation of the wheels.
My idea is a 6 falcon tank tread with 4 pneumatic actuators on the the corners to lift the corners to overcome the scrub on turns making it more agile than other tank tread bots. So turning left you lift the front of the left tread and back of the right tread.
A ton of pushing power and speed as well as better maneuverability than traditional track drives

17

It already happened: :slight_smile:

2 Likes
18

Plugging my own ridiculous creation here, lol

In all seriousness though, probably a West Coast drive or a simple swerve drive.

2 Likes
19

so your saying using a real CVT would be a far better solution?

20

Swerve and Mecanum. Swervanum

Swank Drive. Swerve tank