Robostangs 548 2016 Offseason Swerve - V01 Sidestep

[BigHorse]

http://i.imgur.com/L7OYOid.jpg http://i.imgur.com/2i5SYvX.jpghttp://i.imgur.com/hd7R2q8.jpg


For full sized images:
http://i.imgur.com/6yOe0tc.jpg
http://i.imgur.com/QY07mmP.jpg
http://i.imgur.com/vpKNfSt.jpg


Please enjoy a look at 548's second offseason Chassis! It's a swerve this time!

Features include

Four Independent Swerve Modules
Belt Driven Four CIM Drivetrain
Versaplanetary Integrated Encoders for Steering
Fully CNC Routed 5052-H32 Aluminum Body


If you want to see this fella in action check out our Offseason in Three Minutes video
https://www.youtube.com/watch?v=xE3RIf0Xjfs

Link to the Solidworks CAD files here
https://grabcad.com/library/548-swerve-v01-sidestep-1

[Ari423]

Looks nice! Some questions:

How much does it weigh for the full chassis and per module?
What is the chain shown in the video used for? (I don't see any on the robot)
How are your drivers liking it over WCD?
Would you consider using this in season this year?

[jnicho15]

Looks nice! Some questions:

What is the chain shown in the video used for? (I don't see any on the robot)


Keep in mind that both our early-offseason rocker and this swerve are in the video.

[gerthworm]

Awesome stuff! Did you post the software for it anywhere?

[orangelight]

Awesome stuff! Did you post the software for it anywhere?

The swerve code will be posted soon, I still need to clean it up.

[BigHorse]

Looks nice! Some questions:

How much does it weigh for the full chassis and per module?
What is the chain shown in the video used for? (I don't see any on the robot)
How are your drivers liking it over WCD?
Would you consider using this in season this year?



The whole chassis weighs roughly 60 pounds according to Solidwork's mass calculator. I am willing to bet it weighs a few more pounds than that.
Each module is just shy of 7.5 pounds
The chain is from our other offseason robot.
The drivers love it. It is a blast to drive, and people at events love seeing it run.
While it is significantly more mobile than our traditional 6 wheel rocker design, it takes that much more maintenance. It is definitely on the table now that we know we can build one. If the game needs something fast and agile, who knows?:D

[Chris Hapstack]

How did you decide on the gearing for your steering motors? It looks like the BAGs are geared at 100:1 (assuming both VP stages are 10:1), which would put you at 130 rpm (unloaded). Is that about standard these days for swerve steering? Never having done a swerve drive myself, I'm curious how you would decide on a target rotational speed. I guess iterative testing would be a good place to start - let the drivers try out different steering speeds and find their sweet spot. My only real reference is the AndyMark Swerve & Steer module which steers at 90 rpm.

Also, when it comes to motor selection for swerve steering, how do you model the steering load? I would be inclined to call it a rotary mechanism, with the arm length being half the width of the wheel and the arm load being the friction force on the wheel. I plugged that into JVN's calculator and got the following (http://i.imgur.com/H5fKKc1.png). Does that seem like a reasonable way to model it? It seems like a pretty low loading, so I'm inclined to think that friction in the swerve module itself actually plays a pretty big role (potentially bigger than the wheel/ground friction).

However, if this model seems reasonable, then could you potentially get away with using an AndyMark snow blower motor geared at 1:1? (JVN here (http://i.imgur.com/31Dz8Qi.png)). That would give you 100 rpm steering in a lighter and much cheaper package (albeit without an easily-integrated VP encoder). Experience makes me hesitant to use a built-in worm gear motor for any mechanism that runs the whole match, but then again I've only used window motors before. I suppose the snow blower motors are designed to run continuously, right? At least a snow blower itself runs continuously while a car window doesn't...

[SerpentEagle]

Fantastic!

[Bennett548]

The steering motor gearboxes are 50:1. They do spin very fast, so it would make sense to gear them a little lower to save some battery energy. We have not noticed the motors getting hot, so that is good.

The real test would be trying to steer while being pushed. I know that would take much more torque. We are using needle thrust bearings in the module pivot, so that should help somewhat.

Your model leaves out the friction force in the pivot bushing or bearing. You should also consider the extra toque that would be needed if the robot is getting pushed:

pushing force*height from pivot to ground*pivot bearing radius*pivot bearing coefficient of friction/pivot bearing radius/2= turning torque

N*m*m/m= N*m (unit sanity check)

The divide by 2 is assuming that you have 4 modules, but only 2 take the load because 2 are somewhat lifted by the pushing robot.

1640 did use a window motor for steering, but they only did it for their first generation pivot drive. (http://team1640.com/wiki/index.php?title=DEWBOT_VI_Drive_Train) If you do use them make sure to remove the backdrive locking pins. There is info about that on their webpage, but I can't find it just now.

[Drewit]

1640 did use a window motor for steering, but they only did it for their first generation pivot drive. (http://team1640.com/wiki/index.php?title=DEWBOT_VI_Drive_Train) If you do use them make sure to remove the backdrive locking pins. There is info about that on their webpage, but I can't find it just now.

Here is the page on our site (http://team1640.com/wiki/index.php?title=Nisso-Denko_(Window)_Motor_Locking_Pins) about removing the locking pins on the window motors.

[s_forbes]

I like how all of the components for each module are mounted to a single plate, it makes for a nice simple assembly. The chassis is built in a great way to support each module too. Surprising lack of carbon fiber though!

What kind of control methods have your drivers experimented with, and what do they prefer?

Here is the page on our site (http://team1640.com/wiki/index.php?title=Nisso-Denko_(Window)_Motor_Locking_Pins) about removing the locking pins on the window motors.

Side note: I love your team's website, it needs to be posted more often.

[jnicho15]

Surprising lack of carbon fiber though!

After a dreadful experience last season with our arm mechanism (where we had a four-bar linkage clamp onto carbon tube which always slipped), our new captain plans to avoid carbon in most all cases except if weight is insanely important (like canburgalars) or low-shear loading applications.

[Chris Hapstack]

Your model leaves out the friction force in the pivot bushing or bearing. You should also consider the extra toque that would be needed if the robot is getting pushed:

pushing force*height from pivot to ground*pivot bearing radius*pivot bearing coefficient of friction/pivot bearing radius/2= turning torque


That's a great point - thanks a bunch! I knew a pushing robot would increase the necessary turning torque, but couldn't figure out how to account for it. I like this model a lot.

[Zeb340]

Where did you get the bearings that connect your swerve module to that (I'm assuming) quarter plate? Also are you satisfied with only one bearing attaching the module or because you used one bearing does it wobble?

[Bennett548]

Where did you get the bearings that connect your swerve module to that (I'm assuming) quarter plate? Also are you satisfied with only one bearing attaching the module or because you used one bearing does it wobble?

The thrust bearings are from mcmaster carr: 5909k41 (https://www.mcmaster.com/#5909k41/=15jh4nt)

The plate is 1/4" thick.

There are actually two thrust bearings and a plastic sleeve that make up the pivot bearing. The 1/4" plate is sandwiched between two thrust bearing assemblies. The assembly was shimmed to give the thrust bearings a pre-load. This removes the wobble. The plastic sleeve bushing keeps the "spool" of the pivot from contacting the aluminum plate. I'll ask one of the students to put up a screenshot of the cross section to clarify.

[orangelight]

You can find the code below
https://github.com/Robostangs/SideStep

[Conor Ryan]

Can you walk us through your manufacturing process? From the video it looked pretty cool.

[BigHorse]

Can you walk us through your manufacturing process? From the video it looked pretty cool.

After we finished an extensive CAD model we started the fabrication process. Usually, we travel to GM's Milford Proving Grounds to use their waterjet, milling machines, and lathes, but for this particular chassis we routed all the aluminum on our new Velox 5050 CNC router. We did all of the turning and milling in our robotics workshop in our school. We bought all of our stock from onlinemetals.com. (The cool looking shapecutter in the video was a plasma cutter. This summer we created a relationship with our local community college and they were nice enough to let us make use of their new plasma cutter. We only ended up using it on our rocker chassis.)

We started with the parts to make a single module so that our programmers could have some time with it. We routed the plates from 0.25 6061 aluminum sheet. This was some of our first work on the router so it was a learning process in terms of feeds and speeds. :yikes: While some members routed parts, others were milling and turning. We assembled the first module, tested it out, and made some slight improvements like access to gears and faster toolpaths. After that it was just repeating the same process untill we had 4 modules. For the 0.125 5052 parts we were able to make a toolpath that cut all of them at once, and that saved a lot of time. :D We have a relatively old brake that still works pretty well so thats how we created flanges on the 0.125 material. We used a lexan bellypan so tearing through that on the router was fun. :D After we cut all the parts, it was all just assembly and programming.

It was a lot of work, mistakes, and learning, but I think all in all it turned out pretty well.

[Bennett548]

http://i.imgur.com/wCQkiYm.jpg

Cross section showing roller thrust bearings.