dynamic balancing

[kitestrings]

Has anyone here ever tried dynamically balancing their alternator, or (what I'm most interest in) their blades/rotor?

I've really only seen things like tire balancers, first hand, but I've seen videos of other instruments and methods.  One I've seen, uses a strobe with a "simulated advance".  If I understand it right, they freeze-frame at some synchronous speed, then advance or retard the strobe to simulate a rotation relative to this speed, and somehow index the circumference against measure movement or vibration, to determine where to add weight... or remove material.  These applications always seem to be at 1,500, 2,000 or higher rpms.  Can this work with something that turns say 150 to upwards of maybe 300 rpm?  Are there other methods?

Thanks,

~ks

[joestue]

personally i don't think you need to because of how thin the disk is.

compare to say an electric motor where the length of the rotor is 2 or even 3 times the diameter, and so you could have a static balance, (doesn't matter which end of the rotor you add weight to), but dynamically it does. -and you may not notice until a standard 3hp 3600 rpm motor hits 7200 or higher.

where as for a propeller, its a thin disk. which side of the propeller you bolt the weights to probably doesn't matter.

i did some googling and it looks like what aircraft companies are calling dynamic balancing, just means balancing the prop to minimize vibration once its bolted to the engine (so say the spindle has both some run out, and axial bend). where as a static balance is performed when you spin the prop on a spindle with zero run out or axial tilt.

[mbouwer]

In the workshop you can balance the rotor as best as possible.
I think that there are many influences for imbalance on top of the mast, for example moisture and ice.

[Adriaan Kragten]

One of my first 4-pole PM-generators made from an asynchronous motor was balanced dynamically. For use in a wind turbine, the rotational speed is very low and then balancing isn't necessary. But it was sceduled to use it in a hybrid electric car where it would be driven by a small combustion engine at a rotational speed of about 3000 rpm and then balancing is important. I brought the armature to a company specialised in balancing and this company had a test rig for dynamic balancing. They drilled some holes at both sides of the armature. But the car project was cancelled and the final winding wasn't laid. This generator has never been tested at 3000 rpm. Similar ones with a standard 230/400 V winding have been tested unloaded up to 1500 rpm and loaded up to 1000 rpm without balancing of the armature and this gave no vibration problems (see reports KD 78 and KD 200). The construction of this type of generator is expained in report KD 341 and KD 503. As the grooves in the armature are milled very accurately and as all magnets have the same mass, almost no imbalance is created by transforming a short-circuit armature into a PM-armature.

[bigrockcandymountain]

What do you do now for balancing kitestrings?
The airplane propeller shops do dynamic balancing, and it is supposed to make a night and day difference in felt vibrations.  I don't have any first hand experience with that though.  I'm also not sure what sensors they use. 

I was pretty careful carving my last blades that they all weighed the same amount, and balanced at the same point. Tracking of the tips was good too.  I did a flat static balance, and they seem very acceptable.  Not perfect, but good enough that i don't worry about it. 

My last ones were pretty terrible, and i always say " happiness is relative"  so that could be a factor.

I hope you are having a good winter with lots of sun and wind. 

[Adriaan Kragten]

Balancing of mass imbalance may not be enough to prevent vibrations. There can also be aerodynamic imbalance if all three blades are not exactly the same and have exactly the same blade angle at the same radius. A little difference in lift near the blade tip gives a large difference in the bending moment caused by the thrust. As this moment difference is rotating with the blade, it produces a vibration with a frequency equal to the rotational speed in rev/s. An advantage of constant chord blades is that it is much easier to make all blades the same.

[dnix71]

Adriian is right. I used to pick up and repair box fans.  It's hot and humid here in south Florida, so a fan is a nice supplement to a/c.
Big box fans are difficult to spin balance if they are ever dropped or the blades bent. The same chord is difficult to do by hand. It was usually easier to make each blade push more air so the fan spun slower. If you have a strobe light (old school florescent tube work fine) you can see which blade is causing the wobble.
Just put a number or mark on the front of the blade and let the motor spin up. Adjust the twist until all 3 blades travel in the same circle.

[kitestrings]

Thanks for the replies.

Let me start with what we do now...

I've tried as much as possible to carve the blades in stages, where whatever I do to one, I do to them all.  This includes the removal of larger areas of material of course, but even planing, sanding and finish.  I routinely weighed them.  I ensure equal distances between the blade tips before drilling.  My rotor & hub is too large to suspend horizontally, as some have done.


Until recently, we had always used a similar static balance method:  Accurately weigh each blade, then find the center of gravity by balancing on a knife's edge.  Calculate a moment for each.  Add weight (or remove material) to bring those moments to as close as practically.  When we refinished this set in '20, I was able to get them within about 1.5% with this method.  My buddy Neil swears by this method and says they used it on some comparably much larger (40-60 kW) turbines.


We had the alternator down that year, so I went one step further, which was to assemble the rotor on a test stand and balance to where no one blade was seeking the down position.  This was time consuming, but actually worked very well.  The set we have running is better than I've been able to do in the past.




We also measure and shim for "tracking", locked in yaw, and measuring to a common point on the blade tip to the tower.

So, what is broken?  Nothing really, except I know I will be re-carving at least one, possible a set, next summer due to the water intrusion (documented awhile back).  We do have a resonant frequency just above cut-in - roughly 100-110 rpm - that is a bit worse since the temporary repairs to this blade.

I read somewhere IIRC that the military does dynamic balancing for anything over 150 rpm.  Most of the applications I've seen are much higher speed.

This video got me thinking perhaps it could be done at a lower speed, and perhaps even with more basic equipment.  It's pretty cool:
https://www.youtube.com/watch?v=F5UcdjUsuJs

Plenty snow & cold here bigrock.  High winds today, the turbine is cranking.  Very little sun today.

[topspeed]

Awesome blades.

What is the thickness on percentage wise in the blade tips ?

[kitestrings]

Hi topspeed.  This set is about 2.5" thick at the root and .75" at the tip, so about a third if I'm understanding you question.

We largely followed Hugh's blade calculator, with some initial calc's from Alton Moore's.  Nearest the hub you reach some diminishing returns for your efforts; an area that is not contributing much to the overall output/efficiency.  I believe ours are about (.75" x 5") and under 2 degrees at the tip and (2.25" x 14") 15 degrees at the first 2-3 stations closest to the hub.

[topspeed]

Thickness is chord lenght vs max thickness.

[kitestrings]

The thickness to chord ratio at the tip is (.75"/5)= .15