967 Off Season Project

[404'd]

Team 967 has been working very hard on a prototype and will be posting teasers of our new drivetrain over the next few weeks. The sheet metal frame is all CNC laser cut by a local manufacturing company, DAD Manufacturing, with parts donated by Master Tool Mfg. We hope to be able to take this drive along with an improved gatherer and shooter to off season events this year. And for all wondering, this is not just an ordinary 8 wheel. Any guesses are welcome.
#1:
http://farm8.staticflickr.com/7065/6996900846_a7658971b8_c.jpghttp://farm8.staticflickr.com/7245/7142988693_0d0d1eb140_c.jpg

[CalTran]

this is not just an ordinary 8 wheel. Any guesses are welcome.

I imagine that there is more than meets the eye with this chassis besides it being an 8 wheel, 4 center IFI w/ Blue Nitrile and 4 outer omnis? I don't see much for shifting, though I don't have much experience with custom shifters. (Only have experience with AM Supers)
I see a rectangular hole in the back that I'm going to say is for a battery. And given there are no openings in your chassis, I'd say that you're designing an over the bumper collection system.

Anyways, that CNC laser job looks amazing! Competing against the Iron Lions is always a pleasure, and I look forward to next year! Perhaps we can help you scout again? (Well, more you helping us! :D)

[Andrew Lawrence]

4 center IFI w/ Blue Nitrile and 4 outer omnis?

Looks more like AM Performance wheels w/ Blue Nitrile.

[CalTran]

Looks more like AM Performance wheels w/ Blue Nitrile.

Ah. My bad.

[MichaelBick]

It has drop down casters/omnis in the back. Though with that setup, it would seem as though the robot would be lifted onto the casters/omnis in the back and the front omni wheels. Which would mean you could very easily be pushed.

[PAR_WIG1350]

The cutout in the belly pan could be for a balancing mechanism like a stinger.

[MichaelBick]

The cutout in the belly pan could be for a balancing mechanism like a stinger.

No, it is too big. Actually you can see a drop down caster/omni assembly on the wildstang iteration thread.

http://www.chiefdelphi.com/forums/showthread.php?threadid=105969

[Akash Rastogi]

This is some really solid sheetmetal work. Looks beautiful.

What is the drop on the center set of wheels? What reduction are you running in that gearbox and the overall reduction?

Looks great! I think you could simplify the frame rails and bit for easier assembly but other than that this is great.

[Andrew Zeller]

Awesome drivetrain. What software are you using? What is the best way to model the cutout pattern and have it stop at a specified edge on the bellypan?

[Gray Adams]

Awesome drivetrain. What software are you using? What is the best way to model the cutout pattern and have it stop at a specified edge on the bellypan?

I'm not sure what they do, but it works pretty well to make an oversize diamond pattern and then just use an extrusion to make the border of the sheet. The same applies to the cutout or any electronics mounting you might add.

[BrendanB]

*whistles*

That is a really nice chassis! Can't wait to see the final!

[Nemo]

This is some really solid sheetmetal work. Looks beautiful.

What is the drop on the center set of wheels? What reduction are you running in that gearbox and the overall reduction?

Looks great! I think you could simplify the frame rails and bit for easier assembly but other than that this is great.

Thanks! The gearbox is 50:12, and the chains are 32:16. The axles are all level, but the omni wheel radius is around 0.10" smaller than the performance wheels with nitrile tread.

Assembly wasn't too bad, but next time we'll give ourselves some breathing room in certain places to make it easier (reducing time spent grinding...). What would you simplify on the side rails?

[Nemo]

Awesome drivetrain. What software are you using? What is the best way to model the cutout pattern and have it stop at a specified edge on the bellypan?

This is done in Inventor. Making the diamond pattern basically involves doing the math to figure out exactly how big the diamond needs to be in each direction, depending on the number of diamonds you want and the desired thickness of the material between the holes. I don't know exactly what our student did in this one, but I've just made a couple of diamonds in a sketch, patterned them all the way out in two directions, then cut them all at once. Gray Adams' suggestion sounds good, because it would save you from having to cut the diamonds on the edges in half.

[AdamHeard]

While I disagree with a few of the detailed decisions in this, I can't question the quality or the work ethic that brought this to existence. It's this kind of detailed and quality work in the offseason that will bring your team to the top level.

I love the bellypan, in the future I would try to add material to the bellypan specifically for electronics; it forces a clean layout and lets you mount electronics with the minimal amount of weight.

I think switching to a 2-stage gearbox, then directly driving one of those wheels would clean up the assembly and increase reliability substantially.

The webb of the diamond pattern could be decreased for maximum weight efficiency, but probably isn't too heavy now.

What is the weight of the two outer "rails"? Seems like a lot of material in there that could be pocketed out, especially with the reinforcement bumpers provide.

Why Omni on the outside, you could easily do traction without sacrificing turning speed or ability.

Overall I like it though, it definitely looks robust and quality. Don't take my questions negatively, I just hold drive trains to a ridiculous standard.

[topgun]

I love the bellypan, in the future I would try to add material to the bellypan specifically for electronics; it forces a clean layout and lets you mount electronics with the minimal amount of weight.



What do you mean by this? Specific mounting holes for mounting jaguars, the pdb, etc.?

[Nemo]

While I disagree with a few of the detailed decisions in this, I can't question the quality or the work ethic that brought this to existence. It's this kind of detailed and quality work in the offseason that will bring your team to the top level.

I love the bellypan, in the future I would try to add material to the bellypan specifically for electronics; it forces a clean layout and lets you mount electronics with the minimal amount of weight.

I think switching to a 2-stage gearbox, then directly driving one of those wheels would clean up the assembly and increase reliability substantially.

The webb of the diamond pattern could be decreased for maximum weight efficiency, but probably isn't too heavy now.

What is the weight of the two outer "rails"? Seems like a lot of material in there that could be pocketed out, especially with the reinforcement bumpers provide.

Why Omni on the outside, you could easily do traction without sacrificing turning speed or ability.

Overall I like it though, it definitely looks robust and quality. Don't take my questions negatively, I just hold drive trains to a ridiculous standard.

Thanks for the great reply.

I believe the outer rails are 2.25-2.5 lbs each, based on a pretty similar model that I have on my home PC. Agreed, some pocketing on those would be a good next step.

2 stage + direct drive would be smart, no doubt. We liked one stage because it is so light and simple, it puts the CIM motors a little lower, and it puts the weight right in the middle of the robot. We aren't locked into a particular gearbox, though. Two speed is also an option that we will consider. We'll try to drive this thing hard and see if we can break chains with it, and that might influence the direction we go.

There is a good reason for the omni wheels, but let me wait until the students put out their other teaser. You will probably see what we did from looking at that. We stole that wheel arrangment idea from Team Titanium, by the way. Their 2011 bot has a 2 speed 4WD with four unpowered omnis on the corners. It's a cool drive.

[Gigakaiser]

A wild guess:
A wheel may be dropped through the hole on the frame, tilting the robot on the omni wheels. The dropped wheel is powered and lets the robot strafe.

[404'd]

A wild guess:
A wheel may be dropped through the hole on the frame, tilting the robot on the omni wheels. The dropped wheel is powered and lets the robot strafe.
I didn't think someone would guess the drive correctly this fast. Good guess! Here is a link to a video of an earlier version of this drivetrain. http://www.youtube.com/watch?v=X5rNaMhytZQ
We'll get our CAD model, pictures, and a video of the finished drive posted sometime early next week.

[Gray Adams]

I didn't think someone would guess the drive correctly this fast. Good guess! Here is a link to a video of an earlier version of this drivetrain. http://www.youtube.com/watch?v=X5rNaMhytZQ
We'll get our CAD model, pictures, and a video of the finished drive posted sometime early next week.

Very cool. Why not 3 traction wheels per side though, and the omnis at the other end from the drop down wheel?

Seriously this is cool. It would be such a pan trying to T-bone you.

[gracie.]

i know you guys are still working on it, but how much does it weigh?

[JeremyLansing]

Very cool. Why not 3 traction wheels per side though, and the omnis at the other end from the drop down wheel?

Seriously this is cool. It would be such a pan trying to T-bone you.

We decided to go with the omnis in the back to center the turning circle of the robot in 8 wheel mode, if we had used 3 traction wheels then the robot would have pivoted around the middle set of traction wheels not the center.

[Gray Adams]

We decided to go with the omnis in the back to center the turning circle of the robot in 8 wheel mode, if we had used 3 traction wheels then the robot would have pivoted around the middle set of traction wheels not the center.

If you drop the center 4, then you're only ever riding on 4 wheels at a time. With your weight balanced in the center, you'll only pivot on the middle 4 wheels. All turning as far as traction wheels is concerned is just like a standard 8wd, despite the omnis at the other end.

[Nemo]

i know you guys are still working on it, but how much does it weigh?

Actually, we have it finished and will upload the final pictures this week. With all of the electronics and pneumatics it weighs just under 70 lbs (not including bumpers or battery). There are a few places where we can very easily save weight, but it would be hard to get it much below 65 lbs.

Very cool. Why not 3 traction wheels per side though, and the omnis at the other end from the drop down wheel?

You are correct - that would make the most sense. The omni wheels on the front can't be raised much from the middle traction wheels, because otherwise you have to jack the back end up a lot more to clear all of the traction wheels in 3 wheel mode. With that in mind, we just went with all of the axles at the same height. It didn't occur to any of us to raise only the back wheels as traction wheels until we were putting this robot together.

[404'd]

Here's some finished pictures and a video of the drive. Including a bridge balancing photo demonstrating the weight distribution of the drivetrain.
Video demonstration of strafing and driving:
http://www.youtube.com/watch?v=6gPQ8vcQlVM&list=UUqB8XYfCRqscNRRl5ahrunA

More Pictures on our Flickr stream (http://www.flickr.com/photos/lmrobotics/)

http://farm8.staticflickr.com/7100/7161925530_e3b3661ce2_c.jpg
http://farm6.staticflickr.com/5346/7161988208_3ed8cbe1c8_c.jpg

[Jibsy]

Here's some finished pictures and a video of the drive. Including a bridge balancing photo demonstrating the weight distribution of the drivetrain.
Video demonstration of strafing and driving:
http://www.youtube.com/watch?v=6gPQ8vcQlVM&list=UUqB8XYfCRqscNRRl5ahrunA

More Pictures on our Flickr stream (http://www.flickr.com/photos/lmrobotics/)


I like this a LOT... Seems to achieve strafing ability without all of the negative side effects brought about by other drive systems (weight, complexity, lack of pushing power, etc).


I am curious though.. I think I'm missing how it manages to slide sideways without pivoting when the strafe wheel is so near the back.
Are the rest of the wheels on break mode such that the front omnis don't rotate, and rather just slide sideways on the mini wheels?
Not the best explanation of my question... but maybe you will understand.

[Nemo]

I like this a LOT... Seems to achieve strafing ability without all of the negative side effects brought about by other drive systems (weight, complexity, lack of pushing power, etc).


I am curious though.. I think I'm missing how it manages to slide sideways without pivoting when the strafe wheel is so near the back.
Are the rest of the wheels on break mode such that the front omnis don't rotate, and rather just slide sideways on the mini wheels?
Not the best explanation of my question... but maybe you will understand.

Thanks! When we strafe, in addition to turning the sideways wheel, we also do a turn with the front omni wheels. Those are used to create enough torque to counteract the torque caused by a strafing wheel that is close to the other side of the robot.

[BrendanB]

This is extremely cool! How is the drive system when it is a competition weight? IE 150lbs total with bumpers and a battery?

Very cool design overall to give yourself the lateral movement.

EDIT: Could you post some more pictures of the wheel pod? Is that a custom gearbox or a CIM-sim from AM?

[AdamHeard]

Thanks! When we strafe, in addition to turning the sideways wheel, we also do a turn with the front omni wheels. Those are used to create enough torque to counteract the torque caused by a strafing wheel that is close to the other side of the robot.

I figured this would be necessary for nice performance, about how much power are you applying to counteract the drift?

Are you doing it on some guesstimated value, or using a gyro to hold a heading?

[Nemo]

This is extremely cool! How is the drive system when it is a competition weight? IE 150lbs total with bumpers and a battery?

Very cool design overall to give yourself the lateral movement.

EDIT: Could you post some more pictures of the wheel pod? Is that a custom gearbox or a CIM-sim from AM?

We haven't tested this one with full weight yet, but we will. The previous prototype from the video worked well at close to full weight. That will be interesting to see, actually. The gearing is conservatively slow on the 9th wheel, so my guess is that it will strafe pretty similarly at full weight.

We should be able to eventually take some closeups of the actual wheel setup, but for now I can provide some CAD shots of the gearbox module. By the way, I designed the 9th wheel piece, but one of our students can take credit for the rest of the CAD design.

http://farm8.staticflickr.com/7222/7193277600_d76d330033_m.jpg

This is a custom 3 stage box that is basically a Toughbox plus a CIM-Sim. This is not the newest version (it's on a USB drive at school), so some minor differences exist. In particular, I swapped in the 60T gear from the CIM-Sim when AndyMark released that product. I LOVE that gear. The actual gearing is something like 60:15 / 50:14 / 50:16.

The thing is a puzzle to put together. I probably got carried away a bit in my efforts to make it as small as possible; notice the two holes on top of the gearbox for holding the top pivot brackets on - it is necessary to put the screws in there first, then assemble the rest of the gearbox, then tighten them last by poking the allen wrench between gears. It's a good thing the side plates were bent perfectly at the shop, because they need to be to fit together. Maybe for the next version we can think of something a little easier to make.

We thought about pushing the 9th wheel straight down instead of pivoting it off the back rail/bumper, but we never came up with a clean way to do that without adding a bunch of structure and weight in the back.

I figured this would be necessary for nice performance, about how much power are you applying to counteract the drift?

Are you doing it on some guesstimated value, or using a gyro to hold a heading?

We're using both a fudge factor and a gyro. Maybe Josh will comment more on his programming. The programming presents some interesting challenges, to be sure.

It takes more power from the straight facing wheels than I would have expected. On last summer's 7 wheel version, the straight omni's couldn't keep up with the sideways wheel if we had it in low gear. In the new one, the 9th wheel is geared slower and the one speed drive is faster than the low gear we used last summer, so I think the 9th wheel's speed limits our strafing speed now. Josh, feel free to provide better info on that.

Picking the gear ratio for the 9th wheel is an interesting problem. Ideally, it would be nice if the 9th wheel accelerated the robot around in a circle at the same rate that the main drive's omnis accelerate into a turn in the other direction. That would reduce the amount of correcting needed based on the gyro feedback.

[Nemo]

I'd like to call out a few teams that inspired our team and got us interested in trying a project like this.

JVN and 148 released lots of information about their Nonadrive after the 2010 season, including the
CAD model (http://www.chiefdelphi.com/media/papers/2394) and a lengthy build season journal (http://www.chiefdelphi.com/media/papers/2360).

Madison and 488 put out lots of great pictures, videos, and answers to questions on their Octocanum drive (http://www.chiefdelphi.com/forums/showthread.php?t=90493) starting in 2011, with an update in 2012 (http://www.chiefdelphi.com/forums/showthread.php?t=98137&highlight=octocanum+2). We've seen a bunch more octocanums since then, and I think 488 has a lot to do with that. I don't know if they were the first to do it, but the info they put out was awesome and caught a lot of peoples' attention.

Team 1625 put out info on their Lobster Drive (http://www.chiefdelphi.com/forums/showthread.php?t=92650) last year, and I thought it was a mind-blowingly elegant way of turning a 6 wheel drive into a strafing machine without any extra motors.

I'm not saying these are on the first teams to ever conceive of a hybrid drive (I honestly don't know), but these are the ones who put out a whole bunch of information that we could devour and ponder. And they are all awesome teams who make awesome robots. My students and I really appreciate the openness these teams displayed with their designs, and we definitely learned a lot from them!

[jspatz1]

We stole that wheel arrangment idea from Team Titanium, by the way. Their 2011 bot has a 2 speed 4WD with four unpowered omnis on the corners. It's a cool drive.

Actually the omnis were powered.

[Paradox2102]

Where did you buy your gears from?