Flywheel balance update

[ronmar]

I said in my post above that the chalk method dosn't work.  If I can, I am going to edit that part.  The chalk method works pretty darn well once you understand the lag or delay in the chalk mark from the heaviest part of the flywheel. I think I discounted it's value inappropriately as I actually learned quite a bit with the chalk, but forgot that point when I moved on to the dial indicator. Appologies to Mr X.

I did find that once you get the balance pretty close, you are leaving a chalk mark nearly all the way around the wheel and since the energy imparted to the whole assembly is less, the inertial forces that cause the chalk lag are less or at least different so the ammount of chalkmark lag is different.  I found it a little more difficult at that point to determine the exact spot to add weight so was back to trial and error anyway.

The dial indicator worked well for me at that point as the ammount of frame movement fore-aft is directly proportional to the ammount of movement imparted by the imbalance.  If you are down to trial and error anyway, I found it much easier to spin it up and look at the needle movement on the dial indicator and add weight, then to mark chalk, measure chalk, add weight, clean chalk off, and spin back up for the next chalk mark.

How much is good enough?  If you are going to bolt it down, chalk is probably the easiest way to take the edge off the movement and will also get you started adding weight in the appropriate location around that 76+ inches of flywheel.  But since all the hard setup is done to allow the assembly to easilly move to mark with chalk, it is real easy to use a inexpensive dial indicator to refine it even further.  I used a $7.95 Dial Indicator from Harbor Freight, clamped to a piece of steel setting on the floor.  My generator is still setting in the same location on the garage floor minus the pipe rollers.  It has over 15 hours under load on it now over the last several months just setting there.  You can detect a little movement from the torque pulses under load, but it is not enough to overcome the friction and move the 1000# across the garage floor.

Ron 

[M61hops]

Funny thing happened to me today.  When I quit thinking about what I was doing and tried moving the weights by trial and error I had great sucess in just 6 tries and my 6/1 Metro is a completely different machine!  The chalk marks weren't making much sense to me for a few different reasons but it turns out they did identify the heavy spots correctly.  I can spin my motor both directions with a golf cart motor so that cancels out the lag but I was thrown off by the quality of the surface of the paint on the flywheel.  At first there were pits and drips and chips so I sanded the paint off and repainted.  The new paint grabbed the chalk and made it equally difficult to make consistant markings.  Simple trial and error came to the rescue!  I ended up with about 1.5 pounds of weight about 90 degrees from the cast counterweights.  I'm going to use a dial indicator to see if I can improve the balance but it is very close right now just from trial & error.  To start with the balance was out so far that I couldn't use an indicator because it would have gone around 2 or three times so fast I couldn't see the needle!  I made some weights out of retangular box tubing cut in half to make U shaped channel iron and they look a bit crude.  When I get the weight as good as I can I'd like to make something that looks better.   Maybe if I have them chromed...     
Anyway, the fun continues.       Leland

[ronmar]

Yes, the needle is a blur swinging thru multiple rotations at first, but surprisingly enough, you can actually see improvement on the dial indicator when it is making several rotations.  The pie shaped areas formed by the needle stopping at different spots change size as you reduce the movement and the blurred areas that the needle swings thru change shade as the number of times the needle swings thru them get fewer as the movement lessens.  It of course gets real easy once you have the needle movement under 1 revolution. 

As for weights, you want them to be spread out a good distance around the curcumference of the wheel to avoid concentration of forces which could lead to a crack/break in the wheel structure from stress.  I used 1/4" lead fishing weight wire.  The pieces I used are about 18" long placed along the inside of the wheel rim. I soldered 3 pieces together to make a 3/4" X 1/4" strap that I bonded to the inside of the rims with RTV.  They have since been painted over and you can't hardly tell they are there when stopped and they are invisible when running.  Using this method, I didn't have to drill any holes in the wheels.  With the quality of india castings, this part really appealed to me   

Ron