6wd bouncing when turning

[lemiant]

We have a problem. Our robot rocks wildly when turning at high speeds.
This year we designed our robot to have a similar configuration to a kitbot. We are 26" long by 23.5" wide using 6 wheel drive and 4" Andymark white traction wheels driven via belts off a pair of vexpro ball shifters. The center wheels have 1/8 of an inch of drop, enough that our robot can rock slightly when placed on the carpet. Our weight is distributed towards the back of the robot so that the back four wheels are usually touching the ground. Based on everything I know about drivetrains our wheelbase shouldn't have major scrub problems, but obviously it does. Are there any simple fixes we may have overlooked?

Thanks for your help,

- Alex

[Feroz1325]

This could be caused by a flexible frame, in 2010 we had a similar problem when we built a kitbot inspired frame. The frame was not rigid enough (i.e. we did not give it adequate support for the mechanism load) and this caused the frame to flex while turning and caused all of our wheels to touch the floor making the robot "hop" while turning.
If this is indeed the problem, a few stiffener rails or a thicker bellypan may remedy the situation.

[Jeffy]

Flex in the frame can cause some similar issues. If you haven't, a piece of quality 1/2" plywood for a base plate can help alleviate some problems.

edit: Beat me to it^

[lemiant]

We don't have a bellypan at all at the moment, just a frame, so that may indeed be the cause.
Would 1/4" lexan do the job? Or does it need to be stiffer?

[sdcantrell56]

Originally Posted by lemiant

We don't have a bellypan at all at the moment, just a frame, so that may indeed be the cause.
Would 1/4" lexan do the job? Or does it need to be stiffer?

Try a sheet of baltic birch plywood in 1/4 or 3/8" thickness. If you want to spend more money we have become particularly fond of XX grade garolite in 1/8" thickness

[Jeffy]

Originally Posted by sdcantrell56

If you want to spend more money we have become particularly fond of XX grade garolite in 1/8" thickness

I've used XX before in the 1/32" and 1/16" thickness and hated it. It shatters w/o any flex and was ripped apart by the band saw.
I loved what I found with G-9 and G-10 though.


Whatever you choose, remember it is a fiberglass composite, so wear your masks and gloves when you machine it.

[Jonathan Norris]

Originally Posted by lemiant

We don't have a bellypan at all at the moment, just a frame, so that may indeed be the cause.
Would 1/4" lexan do the job? Or does it need to be stiffer?

1/4" Lexan should do the job. Plywood might be lighter (check the density, Lexan is pretty dense). Make sure its attached at a couple points along each side, that should help alot. Some cross bracing in the middle wheel well might help as well.

[sdcantrell56]

I actually meant g-10. Just rechecked the spec sheet and that is what we have run the past 2 seasons.

[Mk.32]

Originally Posted by lemiant

We have a problem. Our robot rocks wildly when turning at high speeds.
This year we designed our robot to have a similar configuration to a kitbot. We are 26" long by 23.5" wide using 6 wheel drive and 4" Andymark white traction wheels driven via belts off a pair of vexpro ball shifters. The center wheels have 1/8 of an inch of drop, enough that our robot can rock slightly when placed on the carpet. Our weight is distributed towards the back of the robot so that the back four wheels are usually touching the ground. Based on everything I know about drivetrains our wheelbase shouldn't have major scrub problems, but obviously it does. Are there any simple fixes we may have overlooked?

Thanks for your help,

- Alex

As said above stiffness is important in a drop wheel chassis. You could also do omins in the corners as a last ditch fix.

And always on CD a photo would help.

[Feroz1325]

1/4 inch lexan will do the trick, however 1/4 plywood will be more that adequate to keep the frame stiff. We did that last year in our sheetmetal chassis and it worked great! If you have access to a waterjet/plasmacutter/laser you may want to consider a pocketted 1/8 (or thinner) aluminum bellypan....keep time in mind as it is already week 6.
Good luck from team 1325!

[Bstep]

Stiffness will be a factor, but as you mentioned in the original post, the back two wheels are the ones in contact with the floor due to CG placement. This will inherently cause scrubbing because a four wheeled drive base cannot turn without a component of movement perpendicular to that of the wheels rotation. The point of the 2 lowered center wheels is to keep weight off of the four wheels on the outside corners of the drive base (the ones that scrub) but since the rear four wheels are in contact, you essentially created a four wheeled drive base within your six wheeled base. This means the scrubbing is unavoidable, you will only be able to reduce the severity of the bumpy turns by making your frame stiffer.

[cmrnpizzo14]

We use flat twin-wall corrugated aluminum for our baseplate every year. It has worked well for us so far although it might be a little bit heavier than some other options.

[lemiant]

Originally Posted by Bstep

Stiffness will be a factor, but as you mentioned in the original post, the back two wheels are the ones in contact with the floor due to CG placement. This will inherently cause scrubbing because a four wheeled drive base cannot turn without a component of movement perpendicular to that of the wheels rotation. The point of the 2 lowered center wheels is to keep weight off of the four wheels on the outside corners of the drive base (the ones that scrub) but since the rear four wheels are in contact, you essentially created a four wheeled drive base within your six wheeled base. This means the scrubbing is unavoidable, you will only be able to reduce the severity of the bumpy turns by making your frame stiffer.

The length to width ratio of either set of four wheels however is quite acceptable. We built a 4 wheel drivetrain with a smaller ratio last year without any scrub problems.

[Arpan]

Can't say that I've personally tried this, but one stiffening idea that's been bouncing around my team for a while is using tensioned wires along each chassis bar.

[Arpan]

Originally Posted by Bstep

Stiffness will be a factor, but as you mentioned in the original post, the back two wheels are the ones in contact with the floor due to CG placement. This will inherently cause scrubbing because a four wheeled drive base cannot turn without a component of movement perpendicular to that of the wheels rotation. The point of the 2 lowered center wheels is to keep weight off of the four wheels on the outside corners of the drive base (the ones that scrub) but since the rear four wheels are in contact, you essentially created a four wheeled drive base within your six wheeled base. This means the scrubbing is unavoidable, you will only be able to reduce the severity of the bumpy turns by making your frame stiffer.

The point of a six wheel drive is in fact to create a smaller four wheel drive turning base while keeping a large wheelbase for stability. This is a common misconception that I've seen - that 6 wheel drive robots turn on two wheels.
They don't - they turn on four, but on a much shorter turning wheelbase than their stable wheelbase. There's a nice whitepage somewhere which describes this.