pic: 6in Butterfly Module With Piston



(Sorry this description didn’t post until now)

Here’s the next phase of the design I’ve been working on.
Thank again for all of the input on the previous iterations.

Explanation for the piston mount:
Several key things drove this design. I wanted to have as many components of the module be done off the same sheet of aluminum in one run of a CNC laser. Thus, eliminating the need for time on a mill or other piece of machinery.

This is why all the parts on the piston mount are either COTS, or 0.25" Aluminum.
I also want this system to be as “plug-and-play” as possible since my team doesn’t have consistent access to a machine shop. Because of this, I added the aluminum support “tabs” above and below the chassis rail which act as a positioning aid for drilling and bolting the vertical part of the piston assembly to the chassis.
This should make using handtools more accurate and prevent slop that could mess with the module actuation. They obviously also add structural support, but aren’t completely necessary.

I put the piston support on the front side of the chassis because it allows the overall assembly to be shorter. It’s a 0.75" stroke and it makes contact with the ground somewhere around 0.25"-0.35" through the stroke.

A few other comments:
-The actual module hasn’t changed much. I still need to add retaining clips to the axles.
-To answer a few of the comments I had previously, I want to stick with belts/pulleys (I’m probably never designing anything with chain again…) because of the ability to rely on COTS components and again, so we do as little machining as possible. Plus, I want the ease of using a hex to drive everything. I’m just not a big fan of using keyways to drive components.
-I don’t plan on using the axles as added standoffs again because of the necessity of the hex to drive the components for the module.

Looks pretty solid!

  1. Is that a 1.5" or 1" stroke on the piston? My gut instinct would be to balance the stroke so that the traction wheel contacts the ground at about 1/3 extension in the case of a flat game field, but this would of course need to be adjusted in case of ramps, etc.
    .
  2. I agree that 1/4" thickness on the piston mounting plate (assuming aluminum?) is pretty much necessary the way you’ve mounted it, but consider what would happen if you switched the mounting plate to the inside of the chassis rail. This would of course require the mounting plate to bend maybe 10-15 degrees outwards to give enough clearance for the piston, but would place it roughly parallel with the piston’s stroke; the result would therefore reduce the bending moment on the mounting plate, and let you switch to a thinner material or carve more weight out of it.
    .
  3. Neat concept for attaching the piston bracket to the chassis; no welding required! Not sure how much benefit you’re getting out of the bottom plate, though, considering how well the rest of it is already reinforced. Perhaps run some FEA, if you’ve got the capability?
    .
  4. [EDIT] Just noticed that the bolt that you’re using to transfer force between the piston and the module is rather small (1/4"?) for the distance that it’s bridging. Even a grade 8 steel bolt might start to bend during a season under FRC conditions; at a minimum, I would reinforce that with aluminum or brass tubes as spacers (yes, that makes them structural; no clearance here, just tighten the heck out of them).
    .
  5. [EDIT 2] The module side plates *aught *
    to be able to handle everything passed to them by the above bolts, but you might also consider filling in the far-left and upper-right corners (or simply increasing the corresponding fillet radii) of that far-left cutout to more directly transfer the force from the piston bolts to the traction wheels. I’d definitely do the far-left corner, at least, just to be safe…
    *][EDIT 3] Plz also post the same set of images with the piston extended; I am trying to decide whether getting rammed from the side in traction mode could damage any part of this assembly. My guess is that the traction end of the module would flex and rest against the chassis rail or the inside edge of the bumpers before anything started permanently deforming, but having pics would help to confirm. I also wonder about retracting to mecanum mode while being pushed sideways; the open bolt heads and/or nuts may catch on the bottom of the chassis rail or bumper, and prevent the piston from completing its stroke until the opponent lets off.

might be nit picking here, but isn’t a butterfly drive omni wheels as opposed to mecanum wheels?

If you put omnis on it then you’re not going to get much utility out of it. Really, what do you gain from switching between traction and omni? You turn a bit better. That’s about it.

However, if you have mecanum wheels, you can switch between having insane push power, to having a fast and holonomic drive. You get more utility out of it this way.

Tell that to the teams that have used a butterfly drive.

Until you add a strafe module in the middle.

I’m not going to claim to be an expert on butterfly drive, but I wouldn’t feel comfortable cantilevering that module out like it’s a WCD, I’d rather support from both sides here. But then again I’m from the east coast…

I think you are looking at this wrong. The advantage of Omni wheels allows you not to be pinned or caught. Look at 33 from 2014 no one could easily stop them. But they didn’t have a ton of pushing power. The traction allows you to play defense a lot better.

Back on topic and not debating the pros/cons of butterfly vs. mecanum, I have to agree with pmangels somewhat that cantilevering the module sounds quite risky.
I think this would have to be paired with some bulletproof bumpers or an outside rail in order to work super well.

That being said, I love what you’re trying to do, and I really wanna see it work.

How are you going to manufacture the bracket for the cylinder? it looks dove tail joint I would use for wood drawers.

It’s just a slot that allows the pieces to slide together. The fasteners hold it together, but the slots allow for a strength increase. No crazy manufacturing requirements other than a square hole and square tab.

See the Battlebot Bite Force’s construction here

That piston mount plate might bend over time, particularly if there are bumpers contacting it (shoving can get them). We had a similar issue with bumpers in 2014, but it depends on where your bumpers are vertically.

<Grumpy Old Man Rant>

It’s a cylinder, not a piston!

<end Rant>

same. However, to be fair to all “piston” people, if the piston wasn’t there, the cylinder wouldn’t work.

Some items to ponder on this design, but it looks like you are making progress!

Pic 1:
-The black arrows show close proximity of the standoff bolts, the stiffness gain is minimal, you may consider spreading them around to increase the stiffness or eliminating a bolt from the pairs.
-The blue arrow shows the force applied through the module into the cylinder, and the red arrows are the resulting vectors. This places a moment indicated by the second red arrow and creates a high stress area. Whether the .25" plate will yield or not would have to be determined for the application, but the long mounting point increases deflection and reduces stiffness. This would be unwanted for turning in a tank drive with traction wheels as less stiffness will increase bounce during turning. Stiffness is required to allow the wheels to break traction to scrub smoothly. There are many ways to increase stiffness, Skywalker gave one good solution:

[/li]
Pic 2:
-.25" material is plenty thick to tap, the nuts circled may be eliminated by tapping into the back plate.

Pic 3:
-The blue arrow shows the restriction the .25" plates on the top on bottom rails provide, however, there is no loading in that direction. The bolts also restrain the part for rotational loads. It appears this is an attempt to use the pre drilled holes for assembly, if so, the plates could be removed afterwards to reduce weight and be used as an assembly fixture only.
-The green circles are suggestions for rivet locations. They are lower weight than bolts and easily hand assembled. The rivet holes in the .25" can be laser cut like you want, the plate clamped into position, then the frame rail drilled, and then riveted.
-The red arrow is a suggestion to move the pulley to the outside of the wheels to reduce the bending moment on the axles.

[i]

[/i]
Do you guys have more information to support these feelings?

Nah, I don’t actually know how to engineer yet, I only just finished my first year in college!

But great post by you being actually objective! :slight_smile:

To go against this, I believe 624 said that they cantilevered their grasshopper modules in 2014 without difficulty. Whether the fact that that was a grasshopper rather than full Octocanum played into that, I am not sure.

Interesting module, I would also like to see how cantilevering it works out. Make sure you test that before the season starts if you plan on having it as a possibility.

Two things,
First, I too am skeptical of cantilevering the module for stability. I have no doubt that the plates would be fine structurally but I could see where your wheels may want to bend the module and possibly loosen some of the bolts as well as not turn smoothly.

Second, I assume the traction wheels are 3 in, am I right?

I am certainly no expert on this kind of drive and the cantilever may perform wonderfully, I would just have to see in running to be in full support of cantilevering the module. I love the idea behind this sort of drive train and I would love to see a video and real life weight once you actually have it working.

Also relevant is which wheel set – traction or omni – was on the pivot axle.