Swerve Design

[safiq10]

So after months of swerve research I finally decided to take a crack at CAD'ing a swerve module. The module currently weighs 4.28 LBS and the original design intent was for it have its main frame made from a 3D printed composite of nylon and carbon fiber (but if that dosen't work out aluminum is cool too )

The design isn't 100% complete and I'm sure it has some flaws in it. So if anyone has suggestion or ideas for improvement I'm welcome to try them!

Here are the images for it

[R.C.]

Originally Posted by safiq10

So after months of swerve research I finally decided to take a crack at CAD'ing a swerve module. The module currently weighs 4.28 LBS and the original design intent was for it have its main frame made from a 3D printed composite of nylon and carbon fiber (but if that dosen't work out aluminum is cool too )

Are you sure the weight is correct (4.28 lbs)?

Off the bat: CIM (2.8 lbs), RS 550 it looks like? (.48 lbs), VP (.76 lbs). That alone is 4.06 lbs... Add the chain your using and your already at the weight mentioned above. I think you might want to go back through and set actual material properties when you get a chance.

[asid61]

The caster box looks fine apart from maybe too little flange at the bottom, but you need more than one plate to support it up top. In the current configuration, the drive shaft will move around like crazy; there has to be another bearing above all that.
You could probably make the cim face upwards, if you wanted to.
There should be a thrust bearing in between the caster box and the main plate, like this. Otherise your sprocket to turn the caster box (or the chain on it) will be sliding across the main plate when you turn.
What are the standoffs on the VP for? Is the motor and AM9015?

As R.C. said, using some material properties will help you get an accurate weight estimate. When you use "Mass Properties" in SolidWorks, there is an override button you can use for COTS components like the CIM.

[Clem1640]

Some thoughts.

First, I am not sure what bears the load between the rotating wheel/cage assembly and the static drive/steering assembly. There needs to be a rotating, load bearing (the robot's weight, in dynamic terms) connection here which appears to be absent.

I see the hex coaxial drive shaft, which is supported by a bearing in the static rectangular plate (and presumably by another bearing in the top of the rotating cage). Is this the only connection element between the rotating and static portions on this swerve module? I do not believe that this will survive any significant side loading.

We've done a little work with swerve, summarized here: http://wiki.team1640.com/index.php?title=Swerve_Central

I agree with earlier critiques that the stated mass appears to be unbelievable, in the literal sense.

[safiq10]

Thank you to everyone that commented! With your advice I went back in and edited the design up a bit! The new module weight 7.10 Lbs.


The castor now has a second ring to keep the drive axle from going everywhere. Underneath the green ring is the thrust bearing to handle the axle load. I plan to make the green ring out of a new filament from Igus, the iglide tribo-filament.



If you have any ideas for improvements or question please feel free to ask! I will answer any question to best of my ability! And thank you to everyone who has responded so far.

[TheHolyHades1]

Slick rendering!

That weight sounds a lot more accurate. You might want to consider flipping the CIM and the rs550 upside down, and then switching their positions in order to lower the center of mass of the module significantly. That would require you to switch the spacers from the CIM to the rs550, which may in fact be better considering that you'll get a more rigid mounting for the CIM, which will be dealing with higher loads.

Also, I'd be interested to learn how this new igus material performs for you, and how it compares to some more standard choices, like delrin or a bronze bushing.

[asid61]

Apart from flipping the motors to lower center of mass, that looks pretty good.
It looks like you have a 1-stage versaplanetary; what's your reduction on that?

EDIT: I mean overall reduction; versaplanetary * sprockets.

[TheHolyHades1]

Originally Posted by asid61

Apart from flipping the motors to lower center of mass, that looks pretty good.
It looks like you have a 1-stage versaplanetary; what's your reduction on that?

Speaking of the versaplanetary, you should try to move that sprocket a little closer to the base of the shaft if possible - the closer it is to the base, the more torque the system can handle, and the better your chance of not breaking anything. See the versaplanetary user guide for more info.

[safiq10]

Originally Posted by asid61

Apart from flipping the motors to lower center of mass, that looks pretty good.
It looks like you have a 1-stage versaplanetary; what's your reduction on that?

EDIT: I mean overall reduction; versaplanetary * sprockets.

The reduction is a 96:1 I think this will be sufficient for testing purposes.

[asid61]

Originally Posted by safiq10

The reduction is a 96:1 I think this will be sufficient for testing purposes.

That sounds like a good speed, you could go even 10-15% faster if you wanted.