pic: My Rendition of Emperor Swerve

swerve
swervedrive

#1



Height: 6.256 inches tall

4 motors per pod
3 drive 2 550s and 1 CIM
1 550 rotation

Two speeds after losses
7.70 FPS Low
15.40 FPS High

Weight currently undetermined estimating 50-60 lbs including all electronics and wiring.

Comments, Questions, Concerns appreciated.

This drive can be FRC legal, I just figured I’d design something a little bit out of the ordinary :slight_smile:

Hopefully I can get renders of the Pods and rotation modules soon.

https://plus.google.com/u/0/photos/101718392754230534136/albums/5771050412732351601

The album containing the rest of the renders of the system is above, I didn’t want to spam CD media


#2

Pretty nutty design using all those motors but other than that it looks like a beauitful detailed model. Would be interesting to see some more details like the pods. The frame looks potentially very flexy which might or might not be an issue


#3

What are you using for wheel speed and steering angle sensors?


#4

There are some tiny details that are missing, And I wanted to do something different, something that wasn’t for competition use, but easily could be converted to something that would compete. So the 550 Gearbox can be taken off and a new belt put in place and it would power the drive train at an FRC legal drive.

Pod renders should be posted later tonight hopefully.

  • Andrew

#5

Encoders are on the underside of the pod, along with the shifting mechanisms and all motors.

I’ve never designed a crab so I assumed an encoder would be the kind of sensor to use in predicting the rotation of the pods.

  • Andrew

#6

Did you have specific part numbers in mind for the wheel speed sensors, and the steering angle sensors?


#7

This is the encoder I have picked out.


#8

Render Album attached, I didn’t want to Spam CD media.

https://plus.google.com/u/0/photos/101718392754230534136/albums/5771050412732351601

Enjoy

-Andrew


#9

There is a lot to talk about on this, I like the thoroughness in your design.

Probably the largest thing to comment on is the steering setup; Although I am an engineering student, it was pretty simple analysis to show that relatively small bearings (regular roller ones at that) which are very inexpensive are more than robust for the steering setup. If you have an engineering mentor with a mechanical background (or for a rougher analysis, anyone who has done statics or could do a halfway decent Free Body Diagram) they could thoroughly show this concept to you. OR, you could just take my word on it.

We initially looked into large bearing rings and similar setups, and they’re just a lot of parts/complexity for all the ways we had tried to implement it.


#10

I would have to disagree with Adam.

Using larger bearing for the steering in the wheels is not very complex.
Adam’s math may be right and you may be able to get away with regular bearings just fine (973 used swerve this year and if they used them it probably is). But I would say that it is not complex at all to design a swerve using larger bearings. There are several ways to decrease the amount of components it takes in the design.

I would say the dynamic load of thrust bearings is a lot greater, than normal bearings, and I usually fall on the side of it is better safe than sorry, which is why I prefer the larger OD thrust bearings. McMaster-carr sells them for a good price, and even have some neutral file formats available for CAD purposes.


#11

Well what I did for bearings of the steering is a large AL disk that would go inside of a delrin bearing machined for the profile of the AL disk. I’ve seen people do this before with great success. At least I think I have. The pod first started with one bearing on the top and then once I realized the side load would be bad on it I added the bottom circular bearing as well.

-Andrew


#12

I agree with Adam that using a combination of regular roller bearings would be far simpler for this. If you want a cheap thrust bearing we use the $4 Lazy Susan style turntables (6031K31) from mcmaster. We use these in tandem with needle bearings that handle the radial loads.


#13

We are probably saying the same thing and you aren’t realizing it. When I said small bearings, I meant not a large custom disk bearing. Our steering bearings were still relatively large compared to “standard” FRC bearings as we used a pair of 6805-2RS for radial load and a 1.375" Needle Thrust bearing off mcmaster for the axial load (5909K39). Using Metric bearings for radial was the key to keeping the sum weight of the bearings low and still inexpensive.

Our math is definitely correct (couple hundred hours of runtime in abusive conditions), with quite a bit of factor of safety as well. What we did is well on the safe side.


#14

I knew that your math was correct 973 has used the design successfully (and quite fun to watch IMHO).

When you said “standard” FRC I thought that was the normal size, the larger size makes a lot more sense.

However, I still disagree that doing so can make things less complex it all comes down to your ability as a designer, what I find simple others can view as complex.