07-01-2016 01:37

AdamHeard

Re: pic: 2996 Octocanum Prototype

I'd be concerned about the slender shafts that front and rear of the cylinders attach to. They might not have enough stiffness.

07-01-2016 07:55

thedude019

Re: pic: 2996 Octocanum Prototype

This looks great! Is the weight an issue? Ive seen some really heavy octocanum drives before.

07-01-2016 08:08

JoshWilson

Re: pic: 2996 Octocanum Prototype

looks pretty cool, but I'm not exactly sure how that type of drive train is beneficial, could you explain? I've never seen or heard of it before, so I don't know anything about them.

07-01-2016 08:32

Ari423

Re: pic: 2996 Octocanum Prototype

The idea is that you get the best of both worlds: mecanum wheels for maneuverability and traction wheels for pushing. The main downsides are weight and complexity.

07-01-2016 08:48

jnicho15

Re: pic: 2996 Octocanum Prototype

What was your reason for putting the traction wheels on the pivot? In my team's design, we placed the mecanums on the pivot because then, when you retract the traction wheels, you don't need as much cylinder force to hold them off the ground. For traction mode, the mecanums barely need to be off the ground or can be slightly touching- the cylinders don't need to support nearly as much weight. In your design, the cylinders need to hold the whole weight of the robot up off the ground so only the mecanums are touching.

07-01-2016 08:54

Liquid_Science

Re: pic: 2996 Octocanum Prototype

We managed to get the weight right around 54 pounds with everything needed to run the bot (electronics, pnematics, ect.)

07-01-2016 09:04

Liquid_Science

Re: pic: 2996 Octocanum Prototype

The biggest reason is we wanted to use 4in casters and 6in Mecanums. The cylinders we used should be plenty enough to lift even a fully weighted robot, since the boar on each is 1.5" with a 1" stroke, and there are 4 of them. Aslo, we wanted the casters to be the default wheel if something failed. The two wheels can easily be switched around, if we wanted to run 6in casters and 4in mecanums; depends on the game really.

07-01-2016 09:06

JoshWilson

Re: pic: 2996 Octocanum Prototype

Ok, that makes a lot of sense. Thanks for the explanation!

07-01-2016 09:38

aldaeron

Re: pic: 2996 Octocanum Prototype

Have you thought of doing a single larger cylinder between the two modules a la Blue Cheese. Might be less parts and less weight.

What is your gearing for each wheel shaft in the pods from the CIM?

Are you concerned about deflection when the mecanums are down? Seems like a much longer moment arm around the rotation axle. I have wondered this for all the octocanum designs I have seen on CD. It seems like you would want to support the end of the axle that the pod rotates around with a bearing and have an external frame rail running the length. Perhaps the plate between the two pods is enough. Have you tried stacking weight on it to get the bot to 150 pounds?

Love the FIRST logo in the frame =)

-matto-

07-01-2016 09:41

Chris is me

Re: pic: 2996 Octocanum Prototype

Two great reasons to pivot about the traction wheels. First is that traction wheels will see a lot more force in all directions than the mecanums will, especially when being pushed from the side. Pivoting about the traction wheels means these wheels are more rigidly supported by the whole frame. Additionally, mecanum wheels seem to behave better when on a suspension, and the cylinders will serve as a form of a suspension, helping to ensure all of the wheels are touching the ground. The amount of cylinder force for holding one wheel off of the ground instead of the other really shouldn't change in either case.

07-01-2016 12:45

JesseK

Re: pic: 2996 Octocanum Prototype

How does this drive train plan to handle bumpers?

07-01-2016 12:55

T3_1565

Re: pic: 2996 Octocanum Prototype

I really enjoy looking at these types of drive trains. Back in my days these types of drives were never heard of. I made something to try and accomplish what this does.

These are cooler to see people make for sure.

07-01-2016 13:11

JesseK

Re: pic: 2996 Octocanum Prototype

You should provide a link. That type of drive train hasn't been seen since. It was really cool for its time, IMO.

07-01-2016 13:21

nuclearnerd

Re: pic: 2996 Octocanum Prototype

Wow, pretty!

I would really recommend extending the outer plates that join your pivot points all the way out to the front and back plates if you can. I'd be worried about cantilevering those huge assemblies off the side of a pair of bearings that are only < 1" apart! Doing so would also let you support your side bumpers for more of their length (the 2014 rules required support every 8").

07-01-2016 13:35

T3_1565

Re: pic: 2996 Octocanum Prototype

I dont want to hijack the thread lol. Search 1565 Linkage if your interested.

I feel like these types of drives came from what that drive was. Opened up the door for pneumatics manipulating wheels. Before that it was only Swerve Drive.

07-01-2016 13:40

T3_1565

Re: pic: 2996 Octocanum Prototype

I am curious though. Is there a reason that you need 4 cylinders to move the wheels? Can you put a push bar from 1 pod to another and use a wider diameter piston in the middle of the bot? That way you only need 2 pistons and less chance of there being problems with switching wheels?

As I said I'm old and not up to date on these drives lol. There may be a really easy reason why you need them seperate that I'm missing

07-01-2016 14:59

blazingbronco18

Re: pic: 2996 Octocanum Prototype

Looks pretty cool. You guys got a video of it in action?

07-01-2016 17:12

Max Boord

Re: pic: 2996 Octocanum Prototype

What was the reasoning for having the mecanum wheels on the outside? If you get into a head on pushing match you are likely to be tipped off of the colson wheels and onto the mecanums. Also why 6 inch mecanums over 4 inch mecanums and a belt or chain reduction to the colson wheels? It seems like that is the more conventional configuration and would allow for smaller and lighter packaging.

07-01-2016 18:35

Liquid_Science

Re: pic: 2996 Octocanum Prototype

This is just a prototype to test the octocanum theory. Presumably, bumpers would mount to the outside c-channel, and be reinforced at the 2 axles and the piston mounts as well. Bumpers would need some more thought and experimentation though.

07-01-2016 18:37

Liquid_Science

Re: pic: 2996 Octocanum Prototype

Not yet. We still have to get code on the robot and do some testing. But hopefully we will have some video by the end of build season.

07-01-2016 18:39

Liquid_Science

Re: pic: 2996 Octocanum Prototype

Honesty, the main reason is we had 6in mecanums lying around, and this is our first prototype; we just wanted to see what building a more complex drive train was like. But there have been some good points mentioned in the thread, and mine was if something went wrong, we would be able to default to casters.

07-01-2016 18:42

Liquid_Science

Re: pic: 2996 Octocanum Prototype

The biggest reason is it gives us some flexibility to try various other things later. This is just a prototype and will be our test bed for all things related to this drive. It isn't a huge difference to go with 2 or 4, and there are ups and downs for both.