This is the 3rd version of the swerve module. After seeing the Robonauts discussion I added more structure, and better attachments with the base for a rigid design.
Looking good! I’m digging the iterations you’re going through here. Constant design improvement is something to be proud of.
Out of curiosity, do you think you’ll need any sort of tensioner on the lower belt, or will it be robust enough to hold without one?
If you set the center to center distance correctly then you shouldnt need a tensioner, especially considering kevlar reinforced belts essentially do not stretch, especially during a typical first season. The only thing that could be a benefit is some type of idler to effectively increase the wrap around the small pulley. As it is now, with sudden direction changes you could skip the belt
I have seen/stretched Kevlar belting in FIRST. The only belts that I have worked with and not stretched are steel belted Brecoflex belts. From this experience, I would feel safer with some kind of tensioner. Something as simple as a milled slot with a busing mounted idler could do fine.
Thanks
Yeah we have never used belts before. So if idlers are needed I will put them in.
very nice. they look much simpler and lighter than our swerves from 2009!
I wouldn’t say needed, but having more controls over a possible failure point is rarely a bad thing.
Also, I like your method of integrating the frame plate as a pivot surface for the module. Quite slick.
Thanks a ton for your guy’s help.
Ursa engineering and I really appreciate it.
I think we are about ready to start phase two. I am almost done with the total cost of the prototype. Then we will start ordering and punching out the parts. I will be posting pictures of the final system when it is complete.
is that pink piece the only thing supporting the module?
No it is supported there and also on the frame where it is mounted. It has 2 thurst bearings 1 inch apart with a spacer in the middle.
Yeah. I’d say go ahead and put some tensioners in the prototype, and if you find yourself not using them throughout testing, then leave them out if you use swerve for next season.
The other option is to leave 'em out now and see if you feel you need them. This is a good option if you feel really confident about the tensioner design. However, if it’s something you think you might not get right the first time and you’d like to experiment with, give 'em a shot now. Then you can redesign for next season and fix any problems there were.
Keep up the great work!
What thrust bearings are you using in this design? Do you have a link? Thanks!!!
Yeah they are from silverthin.com
They have a ton of bearings there I think I am using a 2 inch id and 2.5 inch od
Random thought regarding this design: I noticed you were using punched and folded sheet metal for your chassis. Have you thought at all about using the same for your module? The precision is doable, and should be cheaper than a pure milled construction. You also might be able to find a middle ground with some parts milled and some punched to cut down on cost.
For example, the top plate of the module milled, with punched sheet pieces for the “yokes” of the module.
Might be something to take a look at for cost reasons.
what do thrust bearings do anyway? I know I should know this, but I don’t.
what specific purpose do they serve in mounting a coaxial swerve module?
A thrust bearing is a bearing that is designed to support both radial load and axial load. Some are designed for more of one than the other. A Thrust bearing: File:Self-aligning-roller-thrust-bearing din728 ex.png - Wikipedia
Here is the link to the bearings we are using.
http://www.silverthin.com/11jsa.htm?part=JSA020
The part number is JSA020.
Thrust bearings are used against an axial force, meaning that the force is coming parallel to the shaft. So for instance with a swerve module the most likely force is going to come from a side hit, which is parallel to the shaft. Thrust bearings increase the strength of the module with the frame so that you can keep driving.
1717 makes coaxial modules out of sheet, look at their design.
Also look at their use and implementation of belting. The belt paths are pretty slick!
http://picasaweb.google.com/RoboticWanderor/FIRST#5313894820442400434