7407 ran the WCP Swerve X flipped corner mount pods this year on our 2022 machine. We created this paper to discuss the modifications we made to the pods throughout the season to fix some issues we ran into.
Here is the paper!
@Dee,
Thanks for sharing these. 131 ran the same modules this season using the âslowâ, 7.8:1, gearing configuration. I want to review our modules because we didnât experience problems #1 and #2 called out in your modifications. It makes me wonder if we have an impending failure or if there is some subtlety in how these are assembled that greatly impacts their use case. This may be something for @R.C to review for future improvements.
I do particularly appreciate your solution to #3. there is a good chance that a 131 robot will borrow a that solution or something similar for next season.
1 Like
Thanks Dan for sharing your gearing info. I do wonder if that had an impact for us, as we went with the âfastestâ gearing available and have already discussed slowing the gearing down in future years.
1 Like
We also ran 7.8:1 and did not observe 1 or 2 at all. This was with all of our competition season + practice on the same set of modules. For 2, Iâd say its observable that the belt is under tensioned, but we never saw anything close to a skip occur.
We did add a frame stiffener as well - but this was mostly due to fear of cable carrier/falling off the bar. Both of which happened and weâre thankful we had them in place. We also used them as an opportunity for intelligent ballast and cut them out of 1/4" steel.
4 Likes
We did something similar to you when bracing the frame. What I really like about the 7407 solution is that they use standoffs to tie into the modules from the bottom frame.
2 Likes
Theres definitely some fertile ground to explore in frame rigidity and overall system tie in vs. some level of flex to the module/system. I can talk myself into either way being better when looking through different lenses.
1 Like
Iâm also surprised by the drive belt slipping. I checked the C-C distance on the CAD model, and itâs bang on for a 55t belt.
Maybe the top plate is flexing (thatâs my best guess), or thereâs too much tolerance stack-up between the mounting holes and the bearings and the shafts. The CAD model does have a really weird eccentricity on one side of the motor pilot hole
2 Likes
Great info, weâve been going through and making our updated mods on our end. Some are inline to improve certain areas that youâve addressed. To start off 1323 has been testing around 12 modules with various configs since last summer. Our practice robot this year saw approx 140-160 driving hours this season. Gone through 4-5 aluminum gear changes tooâŚ
So to address through some of these:
- Currently we have not seen that failure on any of the 3 robots we have running rn. So I do have some questions here:
- How many modules did you see this failure on? 1/2/4?
- How many times did this happen?
- Did you verify the snap ring fully shut on the shaft?
- Was the top bearing a snug fit or loctited in?
- Did you make sure the top cap screw was loctited in?
For this failure to happen, the snap ring isnât doing its job or the top cap came out or both.
Improvements: The reason I ask, we didnât seen this failure at all and we could just replace the snap ring out with a shoulder feature. The only reason we havenât is you lose the ability to swap the bevel gear out without taking the top cap out. Edit: Thinking about this more, might be smarter to just shoulder the bottom section and snap ring the top areaâŚ
- This one has frustrated me here as well, weâve ran this same module throughout the offseason and as the defense robot fully weighted. The drive portion was looser than expected on all modules weâve assembled. While we personally didnât see any failures its something that I personally wasnât happy with.
We think the root of the problem is the slotted feature we did to accept 550/775 motors natively and possibly allowing the falcon motor to flex/rotate towards the big pulley. There is another slot feature on the opposite side but that belt has no issues due to being away from the pulley.
Improvements: We are changing the top plate to remove the slot and allow the option for gears both underneath and above the plate. Belts have gotten a little closer to remove excess gap. We may upgrade the top plate to 7075 as its one of two components that are not 7075 (The wheel is 6061 as well).
All customers who have a Swerve X can get free changed belts & top plate. Should have this out in about 2 months from today. When released send us your order # and any recent order, weâll drop the item in.
- Iâm not following this one exactly, did you have a plate bend? The reason I ask is weâve seen our previous gen swerve have bent plates along with our beta units tearing out at the plate. These plates were all 6061 and semi pocketed. For production this year we changed all plates to 7075 unpocketed.
For reference during hanger testing we (1323) had a moment where we ripped out our backtube because we forgot to bolt together some bracing and some poor design⌠While the tubing had to be scrapped and we had feared that we had bent modules/need to do a full replace they were fine. Attached a picture of our setup this year, used a .090 5052 supporting piece with a metal bellypan on top.
Improvements: We have debated at length about this and will most likely include a corner piece of some sort. Also need to improve manual as many teams thought they should run the module standalone in the corner with no bracing which I donât recommend for any swerve.
Dee/James, I appreciate the feedback and we are making some modifications and new features this year. Teams can expect to see about 8-10 upgrades including but not limited to:
- New top plate
- Option for gears inverted
- Shorter belt
- All steel gear option
- New encoder
- Corner support plates
- Few others weâll wait till we got some items in hand!
16 Likes
R.C,
Chaos 131 does have minor bending in our main plate where it mounts to the chassis rails. The modules are slightly Pidgeon toed in towards the center of the robot. My estimate is that our worst bend is on the order of a few degrees (<5). We found the issue by running our fingers along the outside edge of the main plate. The bend happens right where the module becomes unsupported by the tube. The main plate appears flat across the area with the X-contact steering bearing.
You can see how we support the modules here:
Dan
Did this happen on all 4? I may have to ask for some bottom plates back to look at. On the 1323 robots we built out to the corner and sandwich the bottom of the module to the bottom of a plate.
1 Like
All 4 bottom plates show some level of bending. One or two of them is more severe than the others.
Iâll send you a plate to swap out and most likely build the same test setup.
Currently we make all our test chassis to the examples shown here:
I think the major difference is youâll need something to extend out past the tube and connect to the module and help with supporting the unsupported edge. Iâll also send ya the machined beta corner piece we got to see if that does anything for ya.
Another question are you riveting the bottom plate and then bolt/nut through one surface?
1 Like
@R.C you might consider adding standoffs between your new bracing plate and the main plate the way that @Dee documented in the above report. That would do a lot to reduce any âpigeon toeingâ as reported by @Dan_Karol
5 Likes
@R.C - We just purchased a set of the corner-mount-flipped units to evaluate during the off season. In looking through the example documents you shared (which are awesome, by the way) I really like the âCorner blocksâ you show but it isnât clear to me if they are included in the complete kit or if we need to produce them ourselves. If we need to print them thatâs fine, but can you share the STEP file for them? Also, can you share the CAD model for the example corner gussets you show on Page 6 for the bottom mount configuration and page 8 for the top mount configuration? It would save me a bunch of time from creating my own for the prototype.
Thanks in advance ⌠and thanks again for your support for teams making the migration to swerve!
Evan,
We will be moving manuals to an online portal with the updates that are happening next month. Flipped modules will have option for both gears/belts and corner mounts will come with a support plate. Users that have bought will have an option to get the modified/updated top plate & bottom. Please note that the corner & flipped options will be changing in the next 1-2 months.
Current orders pending will get the option before shipment of what theyâd like (gear vs belt etcâŚ). We donât include the 3dprint corner piece but there are other parts/changes that will be included.
1 Like