Presenting MK4T, The latest innovation in FRC Drivetrains

Have you ever asked yourself, “How would an MK4 Swerve handle a terrain game? Could it scale obstacles? Or is it doomed to fail at a 2” bump in the ground?”

No? Well, I am pleased to introduce my latest creation, MK4T Twerve. While the original motivation was literally “it would be funny”, the real purpose is quite serious: improving the performance of traditional swerve in terrain games.

Onshape Link

Key Details

As mentioned, the objective of the MK4 Twerve Drive is to be able to handle tough terrain games as efficiently as possible. However, to prevent the design from becoming too complex or inaccessible, I made it a goal to make the design as conservative as possible, hoping to ensure that teams could easily adapt this design into their MK swerve module without needing insanely specialized components or extensive modifications.

Everything above the wheel is the same as the MK4i module, including the shoulder bolts and bearings below the wheel (you just need 2 extra shoulder bolts and 4 more bearings). This means that teams who already have MK4 line modules will have a smoother time integrating Twerve.

  • Size comparison

|

  • Weight: 2.6lbs heavier than the same configuration of regular mk4i (6.0 w/ NEO)

  • Designed to be manufactured using a 3-axis CNC mill with tool changer (7475 has a Tormach 1100m) (and a lathe for the billet wheel)

  • Only 8 major custom parts, Two printed pulleys, two-piece billet pulley, two forks, skid plate, belt (+ handful of minor custom printed spacers)

  • Keeps things simple by using the same power transmission system. No additional modes of power transmission are introduced.

  • Uses a common size stock for all parts, all parts are machined from 1/2 aluminum plate stock (1/8 for the bearing spacer)

  • Drives at 80% the speed of the same gearing on regular 4inch wheel setup

Why Twerve?

Standard swerve designs with a 4” wheel typically achieve less than 15 degrees of climb angle on a 1.5” bump, which makes them less than ideal for terrain.

MK4T solves this problem, achieving a climb angle of 45 degrees without compromising the design of the standard swerve module.

Here’s a comparison of the climb angles:

4" climb angle

Twerve climb angle

Comparably, to get the same climb angle at the same height, you would need a 10" wheel to even come close.

10" Wheel compare

Design Variations

MK4T Twerve offers 4 distinct configurations, each tailored to different preferences:

Regular On-center Configuration

This is the latest and most improved version of the Twerve which focuses on giving you the ability to steer the module, which makes it a better choice for general driving.

Pros:
  • Programs and drives the exact same as regular swerve

  • Less mechanical complexity

  • Reuses the same drive bearings and shaft spacer as mk4

Cons:

  • Requires high bumper configuration

  • Requires flipped motor swerve (mk4/mk4c/ flipped mk4i)


Regular Off-center Configuration

This is a base version of the Twerve module focused on accomplishing the main goal of improving climbing performance. However it lacked the ability to steer easily as there is a lot of turning scrub.

Details

Pros:

  • Fits inside frame perimeter a lot better

  • Smaller

  • Does Not require flipped motor configuration of mk4i

Cons:

  • Prog team kills you because you can’t steer

  • more complex power transmission system

  • slower drive ratio


Suspension On-center Configuration

This is a design that used both the steering ability of the regular Twerve and the shock absorption benefits of the suspension.

Details

Pros:

  • Can Steer

  • Can still climb obstacles easily without being a programming nightmare

Cons:

  • X contact bearing sees increased forces

  • Incompatible with odometry as your contact point moves


Suspension Off-center Configuration

This is a blended version between the frame benefits of the off-center design with the added suspension, which can absorb shock when you run into an obstacle or fall down after going over an obstacle. However, it makes it significantly more complex.

Details Pros:
  • Compatible with regular mk4i configuration

  • most efficient configuration for pure obstacle climbing

Cons:

  • Cant steer

  • Greatly increased mechanical complexity




I hope that by sharing this I inspire you to build cool things! If you have any questions or comments, please feel free to post here or contact me directly.

shoutout to @AntPoison for helping with the writing and formating

106 Likes

This looks very impressive, I’m going to go try cadding one for max swerve now

17 Likes

If you make a copy of my doc it might be possible to just modify mine to fit maxswerve, idk tho ive never used maxswerve. (Or you can just make your own)

2 Likes

This is quite cool! Would love to see this implemented IRL.

The suspension ones, in particular, are crazy! I see a few minor issues, but the designs are excellent!

So, you mentioned additional forces on the bearing, and I wanted to explore that here.

The standard SDS modules are about 3.635" from the main bearing to the floor.

Based on your models:
5.006" for the fixed centered, 5.855" for the fixed offset, 6.395" for the suspension centered, and 7.822" for the suspension offset.

That’s anywhere between 38% and 115% increased bearing loading. Granted, the suspension ones very likely decrease the instantaneous forces as the gas shock disperses the force to the bearing over a longer period of time, but any side-loads will still see that increased force.

Obviously it’s hard to predict the forces that are seen in FRC, but I would be surprised if these bearings could survive an entire season with 2x the load a standard module has.

9 Likes

Sorry if I missed it, where do you source treaded timing belt similar to what is designed? I have only seen similar products at Andymark for their drives and they all look to be too long for this type of design.

I got a quote and they can make the belts for $21usd /pc

Considering that the same belts were nearly 400usd when quoted from an american company, i think its quite a reasonable price

17 Likes

This comes from a software dude mindset… so a place of ignorance.

Wouldn’t you want the ‘track shape’ to be a symmetric ‘diamond’ otherwise you may have a flat face against an obstacle? Vs the 45 degree?

Also because some swerve code optimize rotation setpoints to prevent ‘>90 degree’ rotations. With the asymmetric shape. Wouldn’t that be a concern?

4 Likes

Have you done any calculations on the steering torque needed for the offset versions? It may not be as hard for the software as you think?

That said, if a future programming team wanted to get really deep, there was some speculation w free years ago that a “caster” swerve could maybe outperform a regular module because it could change directions faster.

9 Likes

The raw kinematic model of swerve does not have that optimization, so it would be as simple as disabling the optimization in whatever library you are using.

1 Like

Well sure…

But would it be an improvement to make it symmetric? Other than the weight / size increase?

Why do we always have to ‘fix it it code’ ?! /s

2 Likes

in theory yes but then you cant run corner biased modules since it cant start inside frame perimeter

3 Likes

I clicked on this expecting decagon wheels and a billet 7071 frame.

This is actually pretty cool

7 Likes

Ah. I did not think of that. Looking back i see what you mean with extending out the frame perimeter.

Did they give you an idea of what the turn-around time was on a custom order?

The quote says 7-10 buisness days after payment, i cant speak to the accuracy of the quote but it seems rather quick all things considered

Correct me if I’m wrong, but the tread resides fully within the bumper zone. And the bumpers should (in most cases) sit directly against the frame. So wouldn’t the front modules would interfere with the bumpers when moving forward?

1 Like

My assumption is that if its a terrain game the rules would allow for high bumper configurations like we saw in stronghold

1 Like

Since 3D printed tread has seen popularity I would be very interested to see if it would last. But I would be very concerned about going ahead with manufacturing in season without off season testing.

3 Likes

Configurable belt for printing, i might add spikes later if i feel like it

2 Likes

If we don’t do anything with this in season we will definitely explore this as an off season project.

10 Likes