# Wheel inflation and vibration

So upon completing a prototype of a shooter, I’ve noticed a few things that raise some red flags. I used the traction tires on the 4" wheels and when spun to high speeds the tire will start to “inflate” similar to what happens with fast RC cars. I don’t see this throwing accuracy off but I do see it shaking the robot apart. This actually causes the wheel to go off balance and shake the robot quite a bit. I was wondering if anyone else has had this problem and that a possible solution is to building the shooter on top of vibration mounts.

Just make sure to use lock nuts on important parts or to tighten your screws as much as possible.

Not sure what you mean by “inflate”. The traction wheels are pretty stiff rubber. We built a prototype with them last night and it seemed to work quite well. However, we have only run it for a couple minutes so far.

Any vibration or wobbling is more likely do to improper support/bearing use on the shafts.

Make sure the wheel is statically balanced before whipping it around at a few hundred RPM. I have not seen pictures so it is tough to say why it is wobbling.

If you have screws and nuts on the flywheel to make more mass, are they the same screws and nuts all the way around? Same lengths of screws will be important as mass differences causes the center of mass to shift away from the axis of rotation.

Is the Vex shaft about a center of rotation really good? A square shaft in a bearing block round hole could make for things to bounce around a bit. Given the normal shaft is 0.125" across, is that shaft being cantelevered to the flywheel allowing it to be wobbly given a small bend in the shaft? In other words, is the shaft held on both ends and without too much distance in between to bend about.

On cars, the mechanic will add little weights to the rim at the points where the wheel need balancing. What do we have to make things nice and dynamically balanced? A drill to take away mass. Screws and nuts at set intervals of size and distance.

See the first few pages of this to describe what is happening.
http://elearning.vtu.ac.in/17/e-Notes/10ME54/Unit4-VVB.pdf

Could you provide a photo? Also, how did you determine that the tires are “inflating”? My kids filmed their shooter in slow-motion with their iPhone, and the wheel was still nothing but a blur (at 120 frames/ sec, I think they said).

I was in school when i made the post so i couldnt upload another video. I originally up loaded a video on a the thread for designs. Here is a second one where i actually am controlling the speed of the flywheel. You cans see its supported on both sides and the collars are positioned so they distribute weight evenly. If you look closely you can see the rubber lift off the wheel which would be the “inflating”. As i mentioned, this happens to RC cars where when the wheel isnt touching the ground, the wheel will begin to “inflate”. And i agree that the rubber is rather stiff but it can happen. I think the biggest problem would be the axel not being perfectly straight. But i dont believe that is the biggest source of vibration.

I think the inflation could be an illusion caused by the vibration at high speeds. I notice that hou do not have a bearing block on the middle c-channel. I think adding one there would increase stability and at least minimize vibrations.

I originally had one there but it was a major source of friction. Ill add it back to see of it makes any difference.

Nice job with the flywheel.

To me, it definitely looks like the tyre is stretching when the rpms are high. There may not be an easy fix for this.
It also looks like the wheel is ever so slightly off true when it first starts rotating, so this may be contributing to vibration?

Just a word of caution to anyone thinking of adding mass to their flywheel, ensure that there is no possibility that any parts could detach during use. This scares me greatly and I would hate to see somebody gets hurt. You only have two eyes and you want to keep both of them!

It certainly looks as though the wheel is expanding, but I also notice that the vertical C-channel in the background is vibrating a lot more than the one in the foreground. Could you connect those two and perhaps try to reduce the vibration somewhat that way? Among several things, you might be hitting the resonance frequency of that vertical C-channel and amplifying the vibrations inherent in your drive.

Nice!

Just a thought on the vibration. I noticed you used a screw in shaft collar next to the flywheel versus a regular set screw. That additional mass on one side only could be a contributor to the vibration. You have unequal mass off the axis of rotation which could make the inertial axis be oh so slightly off. The RPM is very high. Then again, the mass difference of the set screw is a negative bit but maybe not as much as the 8-32 screw head mass sticking out further from the axle.

Did you measure the RPM yet? I see an IME on the one motor but no wire out of it.

Lastly, it might be a good safety idea to use nylocks on the bearing blocks near the flywheel. Anything that could come loose probably will according to Murphy’s law. Those nuts might suddenly fly around. Locked em down as much as possible. (I know it’s just a prototype, but you can’t get new eyes all that easily.)

So the i actually put the screw instead of the set screw because ive had the set screws come undone during testing and fly across the room. I also have them off set in a way where one screw is facing up and one facing down but i might just replace it with spacers to eliminate the use of collars.

To be honest im not an efficient programmer and I would love to learn how to measure the RPM of this. I originally tried putting the shaft encoder on and the speed just ruined it.

Lastly nylocks was next on my list. But the screw thats in it now has a tiny bit of lock tight but even then that can come loose.

Do you mean it ruined it physically? as in: the plastic was destroyed? Or do you mean you couldn’t get a good reading from it? I’ve been dying to know if the red encoders can physically and electronically take these high speeds.

So I put the shaft encoder on the same axle as the flywheel. And could see readings from it when the wheel spun slowly. But when I reached about 75% power it started jumping with the readings and finally when the wheel reached 100% power the encoder stopped reading, started making this wonderful smell and began to lock it self causing the wheel to slow down. So yes the shaft encoders don’t really work unless you have a slow shooter.

Can’t the encoder just be on the low speed end of the system (motor). If you know the exact gear ratio…quite easy to calculate…you will know the rpm of the flywheel. Once power is applied and the gears are meshed and moving forward, the calculated speed should be exact.

For example, if you are driving a 12 tooth gear with a 60 tooth gear hooked to the motor and the motor moves 100 rpms, you know the 12 tooth gear moved 500 rpm. It had too. Slack in the gears is irrelevant as long as the system is moving in the same direction.

This has confused me as to why people are worried about putting encoders on the flywheel shafts…it is unnecessary…isn’t it?

Sorry for the tangential question, but how will this robot shoot the balls? I’m assuming there is more to be built…another wheel? Or a guide plate, so only one flywheel is needed?

Being able to put the encoder on the driven (higher rpm axle) would be nice because it gives you more ticks which makes accurate velocity calculation easier and stuff like that. But, yeah, not necessary

I would say it’s not necessary to put the encoder on the flywheel end. And based upon what I’m hearing from other people, it’s probably impossible. But I’m still concerned about the red encoder experiencing so much wear and tear at the motor shaft just because of the speed and duration of operation for these units. IMEs might last a lot longer, but my kids refuse to use them.

True. But things might get a little more interesting when kids start to set up speed control. There will be some time lag between the response of the motor and the response of the flywheel when controlling speed to a fair degree. It probably won’t matter at all, but I also won’t be surprised if we hear the gear train “chugging” at times (speed overshooting, etc.). Having the encoder far away from the flywheel will probably make speed control systems a little harder to tune. But I’m just guessing.

So imagine the taller bars in the front cut down to line up with the big C-Channel so its out of the way. I only put them there because I didnt want to cut any metal before i knew this worked. At this point the robot is beyond legal but its still a work in progress. Also there will be a curved back plate with roughly a 4" arc length going around the wheel. We plan to do angle adjustments by slowing the motor but later testing will tell us if its possible or not. Lastly I forgot to take a ball home with me to test with. Our couch insisted on taking May off since we do both FRC and Vex so I was allowed take parts home and do some building. Believe it or not, our team owns 1 of the only vex NBN fields and i forgot to take just one ball home

Couches that belong to my team members insist they get time off, too, but from what I can tell my kids never give them a break. They slouch upon their couches night and day, playing video games and goofing off when they should be learning some trigonometry and how to program their gyro sensor. :mad:

hahaha

I can understand their hesitation. We have used both IME’s and quad’s. The IME’s have proven to be touchy at times, but overall they have worked fine. Basically if we can fit a quad encoder, we will do that just to be safe…but we also don’t hesitate to use IME’s when space is limited. (both physical space, and port space)

Ah, I see. I personally don’t think it will be a problem to have the encoder on the motor side of the flywheel drive train. But I have been wrong many many many many times before. Hopefully we will have working prototypes and the time to test them soon!