Hexagon / Octagon frame with Swerve

[jagoldman]

Hi everyone,

Over the summer I have been designing a swerve module and a chassis frame. I've been seeing all the designs and have learned the benefits of a hexagonal or octagonal chassis as apposed to a square chassis. I was wondering if anyone had experience or pointers on using a combination of a swerve drive and a hex or octagon chassis? Additionally, are there any benefits to using 2"-1" tube to 80-20 in the previously mentioned configuration?

Thanks everyone!

the pdf is of my swerve module. if you guys have any comments on that, that would be great too.

[asid61]

There was definitely a thread on a hexagonal swerve chassis a while ago. A google search might bring it up.

2x1 is lighter than 8020, and it's more versatile IMO. 8020 is easier to work with out of the box though.

Swerve module looks good. I have several quesitons for you:

1. Weight and hight of the module? If it's tall enough, you can place a cim facing upwards, which lowers your COG by quite a bit. Weight is often a major issue with swerve drives.

2. Is that a 1/4-20 bolt on the bottom for the wheel? Anything smaller would be dangerous.

3. What is up with the gear on the top? Keep in mind that means that you have to place your turning gearbox very close to the module, and prevents you from flipping the turning module over like the cim. Changing that to a sprocket or pulley would make mounting much easier.

4. What is the total gear ratio from the input to the output? Just wondering. If it's very low, you can swap out the gears going to the wheel for sprockets to get a higher ratio.

5. How are you constraining the bevel gears on the shaft?

6. What is holding the module on its mounting plate? I see a bushing on the top, but what clamps onto the top of the plate? The weight of the robot alone will not keep it on the module.

Apart from those things, I love the design. It looks very easy to put together and take apart, which is very important in many assemblies. It also looks like it limits assembly to just 1 or 2 allen wrenches.

EDIT: Here's the thread:
http://www.chiefdelphi.com/forums/sh...d.php?t=129095
Ignore my posts on chaining stuff together. I did not understand the whole holonomic/ independent motion control with swerves.

[evand4567]

Another concern with the swerve module (speaking from experience here): You might want to consider hex collars with a set screw and divet or some other stronger form of constraining your hex shaft (like roll pins), because the high amount of torque acting on the bevel gears will try to force the shaft and cause the collars to slip. Having the flange on the bearings facing inward helps, but it's still something you should address.

[Lil' Lavery]

The most often cited benefit of hexagonal and octagonal frames is avoiding friction pins/t-bones. Considering that a well programmed/driven swerve drive should already be able to accomplish this, I'm curious as to your reasoning for pursuing such a drivetrain.

[Bryce2471]

It looks like your wheel axle has threaded holes on either end and bolts holding it to the side plates. I highly recommend that you change this to have a single through bolt. It will make removing the wheel much easier, and probably improve the strength of the assembly.

[EricH]

Originally Posted by Lil' Lavery

The most often cited benefit of hexagonal and octagonal frames is avoiding friction pins/t-bones. Considering that a well programmed/driven swerve drive should already be able to accomplish this, I'm curious as to your reasoning for pursuing such a drivetrain.

You might as well ask JVN what the reasoning was for using a nonagonal frame and swerve drive back in '08. I want to say I remember someone on that team saying it was to make it easier to go through traffic, but can't remember who, or if that was actually what they said.

I would consider it as "ease of getting through traffic" myself--you can use the robot as a wedge to split apart defenders (or move defenders away from the wall) if need be .

[Nemo]

Originally Posted by Lil' Lavery

The most often cited benefit of hexagonal and octagonal frames is avoiding friction pins/t-bones. Considering that a well programmed/driven swerve drive should already be able to accomplish this, I'm curious as to your reasoning for pursuing such a drivetrain.

I thought a main benefit of octagons was to make the robot bigger. Allow wider collectors and such.

[Lil' Lavery]

Originally Posted by EricH

You might as well ask JVN what the reasoning was for using a nonagonal frame and swerve drive back in '08. I want to say I remember someone on that team saying it was to make it easier to go through traffic, but can't remember who, or if that was actually what they said.

I would consider it as "ease of getting through traffic" myself--you can use the robot as a wedge to split apart defenders (or move defenders away from the wall) if need be .

Lap runners in 2008, such as 148, required significantly different strategic design than the "general" robot does. The strategic advantages of round-ish, and more importantly, small for 148 are very different than the drivers behind most hexagonal/octagonal drivebases in 2014.

I don't think you'll see much of a wedging effect, given that bumpers exist.

Originally Posted by Nemo

I thought a main benefit of octagons was to make the robot bigger. Allow wider collectors and such.

It does allow more total area, but actually reduces your sidelengths, resulting in smaller under or through-bumper collection mechanisms. Over-bumper size would be determined by the rules of a particular game's maximum size (how far you can exceed bumper perimeter, total cylinder size, etc.).

[BrendanB]

Originally Posted by Nemo

I thought a main benefit of octagons was to make the robot bigger. Allow wider collectors and such.

We used an octagon frame last year to fit our shooter due to our design concept requiring our robot frame to have some extra width in the center. We didn't realize the benefit of getting out of friction pins until later in the season when teams started asking us how our anti-tbone bumpers were working.

You can pickup a little bit of room for widening a collector by doing something similar to this which is what 469 had this past year. If you play around with geometry in the corners you can gain some room but you need to decide if its worth the effort or if its an aspect of your design that needs to be reconsidered.

[Nemo]

Originally Posted by Lil' Lavery

It does allow more total area, but actually reduces your sidelengths, resulting in smaller under or through-bumper collection mechanisms. Over-bumper size would be determined by the rules of a particular game's maximum size (how far you can exceed bumper perimeter, total cylinder size, etc.).

Agreed. I was thinking of something like 1310's double wide disc collector in 2013.

[Travis Schuh]

Originally Posted by Lil' Lavery

I don't think you'll see much of a wedging effect, given that bumpers exist.

We noticed a wedging effect with our robot this year. The geometry of the octagonal bumpers helped us get into the low goal pocket in situations where I don't think a square robot would have succeeded.

Another advantage of having an angled side is that you can adjust your aim when shooting even with someone pushing on your bumper (that is if you aren't doing a turret on top of swerve).