Two RX7s, duh.
As long as it is one FC and one FB.
By design it’s more “balanced” then the wheel(s) it is attached to.
It is REALLY good practice to balance anything that spins this fast as a unit/assembly. We balance our shooter wheel assembly using a go kart wheel balancer ( the type of go karts that do 100+MPH). It’s not so much about safety but about performance and reliability. Bearings don’t go bad, screws don’t get loose, camera images are clearer, quieter, current draw less, etc.
Do you still have it sitting on COTS 1/2 hex? What did you add to balance it?
I didn’t worry about chasing balancing with a single, 2" wide, 4" dia solid rubber wheel in our shooter this year (between bearings 8" apart from each other), with the assumption that we would see too much shaft flex and hex mating floppery for it to matter that much.
One big advantage of these over making them yourself… They are COTS and can be reused year to year. Subject to current robot rules of course.
While I agree that this is really good practice to do system balancing, I am not sure if this would be something that would be within the capabilities of most teams.
If you don’t have access to a wheel balancer, or the flywheel system geometry does not lend itself to using a wheel balancer, balancing can be done using an accelerometer and an oscilloscope. Those may be tools that teams may have access to, but would probably be harder to find than a gocart wheel balancer.
If you have an accelerometer, tach signal and oscilloscope, it is possible to do the balancing on the robot (in situ). You can get a tach signal from the motor if you are running 1:1 gearing, but if you are running some other gear ratio between the motor and the flywheel, you would probably need to get a tach signal off the wheel itself by adding a temporary encoder if you don’t already have an encoder on the wheel. The accelerometer can be placed on the robot structure somewhere close to the bearings for the flywheel.
If you feed the tach signal and the accelerometer signal into the oscilloscope, you can trigger the O-scope off the tach signal to get the phase relationship between the tach and the vibration to inform where to apply the balance weights. Then you could use trial and error to see how much decrease in vibration you get for a given balance weight applied and you should be able to get the unbalance acceptably low pretty quickly.
Without a tach, you would need to try a couple of different locations for the balance weight and then use some math to infer the phase orientation of the unbalance to cancel it out. After a few trials, if you are careful, you should be able to balance a wheel this way as well.
Something that might be nice would be if these flywheels incorporated a 10-32 threaded hole on each spoke that could be used to add balance weights for system level balancing. You can then install bolts into these holes to act as the balance weights. But using shorter or longer bolts in each hole or even by adding washers to particular bolts, you can adjust the balance very effectively and you have not modified the flywheel.
But, like I said, this is probably a bit out of reach for most teams.
If you are spinning a wheel at a high rpm, and the imbalance is enough of a problem to deflect the axle, then holding a sharpie close to the wheel surface while spinning will mark the side that needs to be corrected.
EDIT: Neat product by the way, clever bore.
If you are deflecting the shaft, then this would work but that would typically require a lot of imbalance. Once you balance out some of that, you would be hard pressed to continue using this method to get the final bit of balancing done.
However, this method of marking the “high spot” this would also work if you were able to mount the rotor group on relatively flexible support structure while you are spinning it (such as spring mounted pillow blocks). The flexibility of the supports would result in large enough deflections that this marker technique should work. If you do mount it onto a flexible support structure to do the group balancing, you need to make sure you keep the balance speed below the system resonance. Once you pass through the system resonance, the system phase change will affect the results.
We used AndyMark HiGrip this year supported on a Thunderhex shaft on Thunderhex bearings (NOT hex bearings). The balancer is really nothing too special. Originally for go kart wheels, it’s just a static shaft on bearings. The secret sauce is in the super-duper good bearings and a well made axle shaft for said bearings (calibration shaft) and of course, a saddle to hold the assembly. Set it up and watch it swing. We use strategically placed hot glue blobs for weight. Once balanced, the calibration shaft is removed and the stock shaft is installed.
Ours is home made but looks like this:
As far as how needed it is, most teams don’t but if you can, you won’t regret it. Our practice shooter wasn’t balanced and worked fine but exhibited most of the symptoms that I mentioned.
If you don’t have a good way to make a balance setup (or are too lazy) just change the relative index of all of your components until vibration is minimized and call it a day. We did that and eliminated all of the symptoms you listed. We were spinning 6in wheels at 2-3k though, so nothing too crazy.
We have not found the need to balance the flywheels. All of the surfaces expect for the back face are machined in the same set up. This results in them being very dimensionally accurate and therefore quite well balanced.
Everyone knows the fd is the best.
thanks for answering keep up the good work!
OK, so you are just using the pendulum method for balancing? And that was good enough to get the balance tight enough for your shooter wheel. That’s great!
I was dubious as to whether that method would be precise enough for such a small diameter wheel at such high RPMs. But that method is certainly easier than spin balancing.
Forget rotary engines, what you really want is a radial engine!
Dubious, rightfully so you are. I fear that someone will try this method using a wooden dowel and gigantic sealed bearing to balance a 2.5" colson.
“Oh great! look at that! It’s balanced already!”
+1 for FD.
Do you have a public CAD model? It’s easy enough to make “close enough”, but I’d like a clean one for weights and stuff on MKCAD. I couldn’t find it on the website. Thanks
CAD files for the flywheel have been added to the product page.