balancing?

[hvirtane]

What is the proper (practical) way to balance (well) a three blade wind rotor?

The problem is that the rotor blades are made of wood, and their gravity centres are not exactly in the same distances from the roots. (As well their weights differ a bit.) 

- hv

[Flux]

All you can easily do is a static balance so make sure that all the blades track in one plane, if not you introduce a couple.

Once you are sure tracking is correct you can do a static balance on the rotor. If you have free turning bearings and an alternator that doesn't cog you can do it with the machine in a wind free place. See which blade rotates to the lowest point and somehow add a balance weight directly opposite. As you have no blade at this point you can add smaller weights to the lighter blades or you can make some form of balance plane to add weights near the root. If you use weights on the blades they will be quite small but if you use a balance plane at the hub you will need much bigger weights. It really doesn't matter if you use large weights at the hub or small weights near the blade tips the end result is the same.

The thing to watch with adding large balance weights at the hub is that you do it in a way that doesn't introduce a couple. You may have to split the large weight into two smaller ones if you can't get it into the plane off the blade tip rotation. if you add a large weight to the front or back of a fairly thick hub it will produce a couple. Equal weights on the front and back face will produce an effective heavier weight at the centre.

If your bearings are not free or you have an alternator with drag or cogging you will have to mount on a temporary shaft or suspend the prop on a string from its exact center and add weights until it hangs level.

With these later methods you have to be extremely careful that the centre you balance from is the same as in its working position , an error in centre placing of a few mm will completely defeat all your efforts as you have the whole weight of the prop running eccentric.

This is a bit brief, there have been many ways described here but the concept is the same , avoid a couple and achieve static balance.

It may be worth pointing out that the twin disc axial flux alternator itself can be a serious source of couples if the assembly is not perfect. You must balance each magnet rotor on its own and never try to correct any unbalance here in the final prop balance as you will never eliminate a couple without a dynamic balancing machine that can resolve the couple components and angles.

If you get perfect static balance and it still wags its tail you can bet you have a couple somewhere.

Hope this helps, if you have any specific questions then ask again.

Flux

[hvirtane]

Thank you Flux for the reply.

It seems to be that dynamic balancing is not easy.
We have been thinking about this already earlier, too.

-----------------------------------------------------------------------------
Static balancing is not difficult. We know how to that. 
Often it is enough.
-----------------------------------------------------------------------------

Our plan is that we would put the centre of the gravity of the blades similar, too.
But we don't know any really easy and accurate way

That 'couple' thing might be enough, but I really don't know.

with best,

- hv

[B529]

Several ways to balance blades. Here's what has produced good results for me.

Weigh all three blades on an accurate scale. Check with your local office supply if you don't have a good scale.

Weigh some lead to make up the difference for the other blades. I use tire weights, the kind that will brake off like a segmented Hershey bar. Most tire shop will just give them to you.

Take the heaviest blade and find the center of gravity. Clamp a thin piece of steel in a vise, balance the blade on that point to find the center of gravity. Mark that point, lets say it's 24" from the root. Put a mark on the other two blades at 24".

Put the next bade on the balancing point at 24". Slide the determined weight up and down the blade until it balances on the 24" mark. Screw/adhere the weight to the blade. Do not be concerned with where the weight ends up, it's where it needs to be.

[Flux]

If it is not free enough turning to balance in its own bearings this method is a good one. The normal suggestions of weighing blades doesn't work unless you determine that the weights are identical at the c of g.

If you do it this way with no other form of checking then it is vital that you get the tip spacing exact so the blades are at 120 deg. Having the prop mounted on the exact centre also is still vital.

It is in no way essential that the blades are the same weight as it will all come out in the balancing process but it makes life easier if you start with something nearer perfection in the first place.

Dealing with a couple is another issue, best make sure you avoid the points that can introduce one.

Flux

[hvirtane]

B529:

what you told is exactly the same method as we were planning to do.

Any other ideas?

with best,

- hv

[kitestrings]

hv,

I've just finished carving a set for our 15'-ter, so I've been re-reading much of what I had on balancing.  I try to be consistent with what I do to each individual blade and weigh them during the process to see if one is getting measurably off form the other two.

I've had generally pretty good luck starting with the knife/edge approach described.  The last set I re-finished this was all I did, aside from checking tracking, and they were very smooth.

Unfortunately some of the older links no longer work, but here are some I had bookmarked on the subject that do.  May find something of use here.
 
http://fieldlines.com/board/index.php/topic,138463.0.html
http://www.mindchallenger.com/wind/axial13.html
http://www.fieldlines.com/board/index.php/topic,139279.html

Good luck.

~kitestrings

[B529]

B529:

what you told is exactly the same method as we were planning to do.

Any other ideas?

with best,

- hv

hv,

this method has worked well for me, no reason for me to try others, stick with what works. As Flux suggested, get your tip spacing, blade centers and also tracking as perfect as possible. I would do this regardless of balancing method. Once again do not be concerned about where on the blade the weight ends up.

[birdhouse]

i think i balanced my blades in a pretty wacky way, but it seems to work well, as i had my mill free-spooling and it got ripping pretty darn fast with no tail wobble.

obviously first make sure the tips are all in the same plane and tip to tip measurements are spot on.  then, i mark each blade (A, B, and C).  i then create a chart with AB &C in the header.  i put the mill indoors on a test stand and just give the rotor a spin.  i then put a tick mark in the chart under the column of which blade stops at the bottom.  spin again, another tick, spin again, another tick ect. ect. ect. ect. ect. ect. ect.............

you'll start to see patterns in which blade(s) land at the bottom more often.  i then add weight, and do it all over again.  eventually, you should get it dialed in to the point where each blade lands at the bottom as much as the other two.  then you've got it.

downside to this method is is takes lots of time, and is boring, and thus takes lots of beer   

goodluck!
adam

[fabricator]

Depending on the size of the alternator the place to start is by balancing each magnet rotor individually, if you don't you can end up with a funky couple in the alternator itself and that will just compound any imbalance in the blades.

[Capt Slog]

Instead of fastening things to the blades, I balance using adjustable weights.

I have lengths of threaded bar which are mounted opposite to each blade and they have weights which have threaded holes in them and can be turned up or down the bar.  I have various sizes of weight, and each one has a locking nut to hold it in place.

I think I start by having each blade horizontal and adjusting its opposite weight to balance it, but of course, each adjustment affects everything else.  It's not quick, but it is fairly easy.



Not a good photo but you can see of them in the picture.  The anchors for them are bits of aluminium angle.  I've not had one thrown or moved in the 3 years or so that I've been using them.

[kitestrings]

downside to this method is is takes lots of time, and is boring, and thus takes lots of beer   

I usually find my balance getting further off after a few beers, but I like the sound of this...

~kitestrings

[kevbo]

Another way is to suspend the rotor horizontally from it's center point.  It has to be done either indoors or dead calm. You may need some sort of adapter to attach a rope, chain , or cable to the center. It is important that that attachment be accurately centered.  The closer to the plane of the blades (lower) you can put this, the more sensitive it is.  This has the advantage that you can just set weights and screws on top of the blades and attach them at these locations later, and you get feedback quickly vs having to attach them before you can retest with the "spin it" technique, and needing numerous trials to average the error due to friction.  This is how automotive tires were balanced before dynamic balancers became common place.

[neilho]

I like a slight variation on hvirtane, B529 and kitestring's methods. Using the knife edge, find the blade cg distance from the center of rotation, weigh the blade, multiply distance by weight (resulting in moment), compare the 3 blade moments.

This is much like the method used on many larger machines. Imagine trying to static balance three fifty foot long blades, assembled to a hub. Now compare that to handling one blade at a time, without a hub, without assembly. Much easier, not to mention that it simplifies replacing a blade in the field considerably.  And it's the only method that quantifies the balancing. Dangling the rotor assembly or setting it on a point don't tell us how balanced it is or how much more balanced it needs to be. Gougeon Brothers (the folks who designed, developed and built the early wood/epoxy composite blades used on Mod-1, ESI and Enertech machines) arrived at this method and concluded that blade moments within 2% of each other made for adequate balancing. This quantitative comparison isn't possible with any other method.

Neil

[fabricator]

I like a slight variation on hvirtane, B529 and kitestring's methods. Using the knife edge, find the blade cg distance from the center of rotation, weigh the blade, multiply distance by weight (resulting in moment), compare the 3 blade moments.

This is much like the method used on many larger machines. Imagine trying to static balance three fifty foot long blades, assembled to a hub. Now compare that to handling one blade at a time, without a hub, without assembly. Much easier, not to mention that it simplifies replacing a blade in the field considerably.  And it's the only method that quantifies the balancing. Dangling the rotor assembly or setting it on a point don't tell us how balanced it is or how much more balanced it needs to be. Gougeon Brothers (the folks who designed, developed and built the early wood/epoxy composite blades used on Mod-1, ESI and Enertech machines) arrived at this method and concluded that blade moments within 2% of each other made for adequate balancing. This quantitative comparison isn't possible with any other method.

Neil

What would be the procedure if you find large variations in the moments

[neilho]

What would be the procedure if you find large variations in the moments

Much the same as the other methods. Check blade shape, length, etc and correct if necessary (often the case if moments vary a lot), add weight to the blades with the lower moments, recheck the moments, continue till the moments are within tolerance. One can calculate location of the weight if the weight of the weight is known (sorry, couldn't resist)  .  B529's method sounds pretty slick, too.

My personal preference for weighting is to drill a 3/8" hole into the leading edge of the blade, put in the appropriate length of 3/8" round lead and add West system epoxy to bind it together.  After the leading edge tape and paint goes on over it, it's aerodynamic, secure and protected from the weather.

The point of calculating the moments is to stay within tolerance and to have a record for future use. 

Neil

[ghurd]

Instead of fastening things to the blades, I balance using adjustable weights.

I have lengths of threaded bar which are mounted opposite to each blade and they have weights which have threaded holes in them and can be turned up or down the bar.  I have various sizes of weight, and each one has a locking nut to hold it in place.

I think I start by having each blade horizontal and adjusting its opposite weight to balance it, but of course, each adjustment affects everything else.  It's not quick, but it is fairly easy.



Not a good photo but you can see of them in the picture.  The anchors for them are bits of aluminium angle.  I've not had one thrown or moved in the 3 years or so that I've been using them.

If I am not mistaken, that was Zubbly's prefered method of balancing.
G-

PS- Nice to see you back!

[rossw]

If I am not mistaken, that was Zubbly's prefered method of balancing.

I thought Zubbly was a fan of the "pendulum"?

This temporary one of mine was very ugly because it stuck out the front causing a dynamic imbalance, and was ultimately a longer arm with a shorter weight, flipped over to be in the same plane as the blades.

[Flux]

I think Zubbly did prefer the "pendulum" method. As you say it works fine as long as the weight is in the plane of the blade tips. That applies to all methods of adding weights at the root. If you are in another plane you introduce a couple.

For larger machines where the blades can be fixed correctly at exactly 120 deg, especially those using commercial blades, the method of balancing each blade for a fixed moment at the c of g is very convenient, when things get over 10ft diameter it is difficult to find space to handle such things away from wind, it makes a lot of sense to deal with the individual blades.

For smaller props especially where you have wooden blades screwed to a wooden hub I find it easier to let the blades turn out as they do and not bother about their weight, then balance the assembly with weights placed anywhere convenient. With these small props even if you make the individual blades balanced perfectly you have to go to great lengths to space them exactly at 120 deg.

When you are dealing with machine produced steel hubs with machined faces exactly at 120 deg then balancing the individual blades makes more sense especially with fibreglass blades that should have a very accurate root attachment. Wooden blades may still not mount sufficiently accurately unless produced by cnc or a profile coppier.

Flux

[TomW]

I thought Zubbly was a fan of the "pendulum"?

This temporary one of mine was very ugly because it stuck out the front causing a dynamic imbalance, and was ultimately a longer arm with a shorter weight, flipped over to be in the same plane as the blades.

Yep, that was how Zubbly (R.I.P.) showed me how. Mine usually end up bigger and the last ones were flat bar with holes for piling on bolt, nuts and washers. No picture handy of the current ones.

Tom

[neilho]

. If you are in another plane you introduce a couple.

For larger machines where the blades can be fixed correctly at exactly 120 deg, especially those using commercial blades, the method of balancing each blade for a fixed moment at the c of g is very convenient, when things get over 10ft diameter it is difficult to find space to handle such things away from wind, it makes a lot of sense to deal with the individual blades.

For smaller props especially where you have wooden blades screwed to a wooden hub I find it easier to let the blades turn out as they do and not bother about their weight, then balance the assembly with weights placed anywhere convenient. With these small props even if you make the individual blades balanced perfectly you have to go to great lengths to space them exactly at 120 deg.

Flux

Not sure what a couple is, but lit looks like I've got some reading and education (read: fun!) ahead of me.

Most of the blade balancing I've done has been to wood blades sandwiched between individual plates and a flat steel hubplate, held together with bolts. The bolting system clearance won't guarantee 120 degree spacing, but that's easily corrected at final assembly. Tracking and pitching are done between the shaping and finishing processes.

Agreed that balancing wooden blades screwed to a wooden hub is easier and more appropriately done empirically, as a unit.

Neil

[wolfe]

i know this is an older post, but was just wondering what exactly a ''couple'' is.

I haven't seen it mentioned in other balancing posts

[Flux]

I have covered this before but I wouldn't know where to find it. I will keep it simple and brief.

If you have a disc on a shaft and it is out of balance you can correct it with a weight bolted to it or a hole drilled in the heavy bit. This is static balance and you can sort it out with the shaft in bearings or on knife edges. This is what you are normally dealing with with a prop that is large in diameter but thin axially.

Now imagine a shaft with 2 discs separated by a finite distance ( several inches or more). Now drill a hole in each disc but diametrically opposite, if the holes are the same size and at the same radius the thing will still be balanced on bearings or knife edges, it is still statically balanced but when you rotate it you will have an out of balance force on each disc because each disc is not balanced, one end of the shaft will be pulled one way and the other end will be pulled in the other direction. The two opposite forces trying to rotate the shaft about its centre form what is known as a couple.

It will be statically balanced in this case and you can't correct it without rotating it and balancing each disc. You need a machine to know where to put the balance weights except in this very simple case where you can see the individual balance planes. In real life you usually have a combination of static and dynamic unbalance and without a machine it will take forever to correct it by trial and error. although there are a few mathematical ways of doing it with a trial weight it is near impossible without a computer programme.

You can get this problem with badly tracking blades as they rotate in different planes. You can also get it with the magnet rotors and they are spaced wide enough to cause trouble. Always balance each magnet rotor on its own, don't try to balance the assembly.

I don't want to go deeper into this now but I hope you get the idea. When balancing a prop with heavy weights at small radius keep them in the plane of the blade tips and you may need to split the weight into two parts on the front and back hub if you can't.

Flux

[prasadbodas2000]

very perfectly explained by flux in context of a system with shaft and rotors...in simplest form a couple is a pair of (equal and opposite) forces acting on a body but not passing through same point....