FRC488 Driving Over the Bump

[Madison]

I just uploaded some video to our YouTube account showing our chassis traversing the bump.

First test at slow speed:
http://www.youtube.com/watch?v=kBzcQssOPOA

Attempting to turn while going over:
http://www.youtube.com/watch?v=KvvDzz-9Xmw

High speed pass over the bump:
http://www.youtube.com/watch?v=lGYnysfjeS4

A few notes about what the robot experienced:

• The "security mode" button on the gaming adapter was facing upward. The shock of impacting the ground was enough to cause it to press itself and make the gaming adapter lose connection with the driver station.
• The old IFI transmission bridge that we're using to temporarily support our battery deformed plasticly due to impact. It's bowed about 1/8" downward now.
• The fans in the Jaguars appeared to unseat themselves or otherwise be thrown off balance. They are very noticeably hitting their housings now.

[Jones571]

some great information thanks for sharing

[Vikesrock]

Thanks for posting these videos and especially the info about the parts that appeared worse for wear after the tests.

This was with no suspension or dampening mounts for the electronics right?

[Madison]

Originally Posted by Vikesrock

Thanks for posting these videos and especially the info about the parts that appeared worse for wear after the tests.

This was with no suspension or dampening mounts for the electronics right?

The electronics are temporarily mounted directly to a sheet of 1/8" polycarbonate that is coupled to all structural members of the chassis. We will add additional dampening in the final assembly.

Also, here's another high-speed pass:
http://www.youtube.com/watch?v=UihTfb9KXg8

[waialua359]

thanks for sharing.
We are about to test our 8 wheel 8in. drive tonight with a 2 speed shifter.
**dinner break right now**

Update 1 hr. later:

Drives very similar to yours. The front end being slightly heavier was much easier going over the bump. We let gravity do its magic while coming down up until the front wheels make contact with the ground.
Then you can rip it.

And yes, with everything loosely mounted on our prototype bot, some things got loose, but easily fixable for the real thing.
Overall, we like it.

On to our real kicker now that we already prototyped and satisfied with.......

[MrForbes]

Thanks for posting this Madison. We tried ours out and had some difficulty getting over the bump. Having "stuff" such as the framework to hold the bumpers and ball kicker mounted up higher seems to make it more difficult, because the CG moves up and the robot wants to sit at a 45 degree angle when rearmost wheel gets to the ramp.

I suggest you give it a try with more weight on the robot, approximating the final CG position.

[BHOP]

where is the center of gravity location on this bot?

I would be curious to see if changing the CG to one side and going over the bump with that side would prevent the robot from jumping or having the whole front go up in the air.

Thanks for the videos.

[Madison]

Originally Posted by squirrel

Thanks for posting this Madison. We tried ours out and had some difficulty getting over the bump. Having "stuff" such as the framework to hold the bumpers and ball kicker mounted up higher seems to make it more difficult, because the CG moves up and the robot wants to sit at a 45 degree angle when rearmost wheel gets to the ramp.

I suggest you give it a try with more weight on the robot, approximating the final CG position.

Originally Posted by BHOP

where is the center of gravity location on this bot?

I would be curious to see if changing the CG to one side and going over the bump with that side would prevent the robot from jumping or having the whole front go up in the air.

Thanks for the videos.

We're keeping an eye of the CG and doing everything we can to keep it centered. We've found that if we significantly bias it toward one side, it makes the ride up smoother and the trip down more violent as all of the weight comes swinging down toward the ground. We think it's just nicer to our frame and axles if we keep most of the weight away from the ends that hit the ground.

We should, hopefully, have the rest of the torsion box frame together and attached on Tuesday.

Attached is an image showing the CG of my CAD model. It's missing the electronics, but shows the added weight of the bumpers. The electronics will shift CG fore or aft, but probably won't raise it much at all. Ditto for all of the motors driving the remaining mechanisms.

[MrForbes]

Interesting! Do you have a number for how high the CG is now?

When the robot is climbing the ramp, at a 45 degree angle, the height of the CG affects where the balance point is, relative to the 3rd wheel. Having a higher CG moves the center of mass behind that wheel, so the nose does not want to come down as it tries to crest the bump

Anyways, we're hoping that moving our CG foward (or back and driving over the bump backwards) as well as increasing traction on the center wheels will give us the extra umph we need to get over the bump.

Thanks again for posting this info!

[Madison]

Originally Posted by squirrel

Interesting! Do you have a number for how high the CG is now?

When the robot is climbing the ramp, at a 45 degree angle, the height of the CG affects where the balance point is, relative to the 3rd wheel. Having a higher CG moves the center of mass behind that wheel, so the nose does not want to come down as it tries to crest the bump

Anyways, we're hoping that moving our CG foward (or back and driving over the bump backwards) as well as increasing traction on the center wheels will give us the extra umph we need to get over the bump.

Thanks again for posting this info!

The coordinate system is all messed up in my assembly, so it thinks the CG is -x, -y, -z. The CG ends up being a little bit above the finished frame when the bumpers, kicker, and upper frame are removed.

We're seeing the behavior you describe even now. In the video showing the slower trip over the bump, you can see that it doesn't begin to tip until the third wheel crests. We're using 8WD for this reason, actually -- it slightly lowers the CG over 6WD because the ground clearance can be lower and the 2nd wheel impacts the bump before the 1st impacts the ground. It breaks up the shock into two more manageable chunks.

As things stand now, the robot can climb to just before its tipping point and old that position, though I can't imagine when that'd ever be a good idea. It'll be interesting to see if it changes, though.

[dtengineering]

You're a couple days ahead of us, but when we watched the video and looked at the CAD yesterday, there was an immediate "Hey, that looks just like our design." response.

We've got an 8wd, low CG "box bot" in the works, too. Some differences, perhaps... toughbox at 14:88:1 direct driving IFI 6"x1.5" wheels on the centre axles and 6" AM kitbot wheels raised 1/4" on the ends. Mostly baltic birch construction.

We contemplated the "impact" problem and have decided that rather than using suspension or pneumatic tires, we're going to try using foam to isolate the battery and electronics from the frame. It's going to be awesome to see some of these robots "get air" off the bump. I pity the poor machine that gets in the "landing zone".

Jason

Okay... technically, I'm not sure that the machine getting air is going to come off any better than the machine that it lands on... but this is going to be one rough and tumble game when the pressure starts to crank up.

[=Martin=Taylor=]

Wow. That’s really scary. I'm surprised so many people are planning to 'brute force' the bump like that. Even if your CG never goes outside the footprint the sheer momentum of your robot pitching forward may be enough to tip it.

If you left larger gaps between the wheels the transition would be smoother.

[Madison]

Originally Posted by =Martin=Taylor=

Wow. That’s really scary. I'm surprised so many people are planning to 'brute force' the bump like that. Even if your CG never goes outside the footprint the sheer momentum of your robot pitching forward may be enough to tip it.

If you left larger gaps between the wheels the transition would be smoother.

The transition is only potentially smoother if the CG ends up directly in the center of the robot. Otherwise, by moving the center two wheels nearer to the end with a CG that is outside the center, you'll spend more time tipped back at 45* before tipping forward.

[MrForbes]

I am very impressed with the design of the bump. It's a wonderful challenge. I hope we can conquer it!

[Joe Johnson]

I am usually a big fan of 6WD but I think this is the year of the 8WD robot.

Actually (believe it or not), I am designing up a 12WD robot this year -- with 8" pneumatic tires -- I can hardly wait to see this beast in action...

Sadly, the design is still under embargo so for now I will leave the arrangement of 12 8" wheels in a 38"X28" chassis as an exercise for the reader...

...more to come...

Big thanks to Madison & Team 488 for sharing not just the design but the videos as well. Very generous.

You're a better man than me ;-)

Cheers,
Joe J.