12 foot wind turbine failure

[DanB]

Last night we had some fairly tremendous winds.  Tom (the owner of the 12' turbine that we built at a workshop here in Aug.) came by early this morning saying that his blades were torn up.  Pictured above is the machine as we found it when we arrived.  If you look close, you can see that the tail stop (the bit we weld to the tail boom to keep the tail out of the blades) is jammed behind the yaw bearing and the tail is stuck in (actually past) the fully furled position.  Obviously the tail hit the blades.  We figure the whole tail lifted enough on the tail pivot to allow this to happen.  Perhaps we should start bolting the tail bearing down so it can't lift off.  I've never seen this before anyhow.

We lowered the tower, Tom, Rich and George inspect the damage.

Another problem!  Normally we have a large steel hub on each side of the blades so that the wood doesn't crush when we tighten the nuts that hold the blades on.  This time we just used washers on each side of the blades.  After two months, all the nuts were coming loose.  I've seen plenty of Hugh Piggott machines that were put together this way, I've never tried it though - I've always used steel hubs instead of washers.  This was the first time we did things this way and it failed quickly.

So she came back to the shop today.  We'll get a new set of blades and fix the tail so that it can never (knock on wood) hit the blades again!  The other big problem with this machine was simply too much power, we've seen sustained output over 2kW from it with peaks well over 3kW.  Last night, when it failed, the current exceeded that which the meter could record so we have no idea what peak power was last night (it blew out the kWh meter).  It needs to furl earlier so George just cut the alternator off the yaw bearing and we're adding another inch or so to the offset.  As it was, the alternator was off set 7", Im surpised it furled so late actually.  Now we're making it a bit over 8", we'll see if that helps.  Otherwise all seems well, the tower works well, the alternator looks fine with no signs of overheating.  Should be a quick fix, an hour or two of metal work and a new set of blades.  

[imsmooth]

Why don't you use lock-nuts instead of regular nuts so they don't come loose?

[electronbaby]

yes, Nylock nuts work well for me. I have always done my props without a steel plate, just nylocks and washers....or do what we do when we do a commercially available machine, use locktite everywhere. Its makes a lot of sense and prevents failures, and most importantly saves me a trip out to the sites. :-)

[zeusmorg]

 It looks to me like the problem isn't in what was used as far as nuts and washers, but the fact that the wood compressed under the washers.

I do believe a way to make sure the tail doesn't lift should be in order, like a bolt through the top of the tail pivot and maybe a stop that keeps the tail out of the blade area totally.

 Also was there any damage to the coils? any sign of overheating?

[wdyasq]

From a long way off .... and not able to personally inspect. These folks are quite able to analyze and solve problems and will have a better grasp than someone who looks at the not-so-pretty pictures.

First, nylon lock nuts do no good if the wood shrinks or gets compressed locally as WILL happen with washers.

If a cable was used to 'tie the tail on the pivot', it could also be arranged to act as a second safety. If a proper spring and a cable is used it might be able to slow the tail before it impacts the stop. I'll sketch something up on this and mail it to DanB.

I've long thought the tail should naurally furl and a cable under tension should be used to 'activate' the machine.

But - this in not my project and I am a long ways off. As it is not in my back yard, I can just yell.

Ron

[electronbaby]

I suppose the wood shrinkage depends on your climate. I live in a coastal environment, although not exactly on the ocean, you can see the Atlantic from the tower top (aprox 1.5 miles away). I have never seen substantial shrinkage on my props (made of laminated cedar) and I have been flying machines for years. Dan lives in a much dryer climate I think, so I would think it would be much less, but I may be wrong. I use pretty wide washers under my lug nuts too, maybe 1.5" or maybe 2" in diameter, so this im sure helps, rather than hurts the situation. The washers shown in Dans pic are not very wide. Split washers also, I would think, would try to loosen the nut if it was not torqued correctly in the first place, specifically in a high vibration environment such as this, and specifically if it was stainless steel. The soft wood of the prop and use of a narrow washer would allow you to torque to much less than if you were to spread the force out over the surface of the wood more evenly with a wide washer, or best yet, a steel plate.

I like your idea regarding the furling cable. It would be nice to have the machine furl automatically if either something goes wrong, of the cable breaks.

[hiker]

tail looks fine--whats it made of?

if made from wood why isnt it torn up?

does the pole have any marks on it?

 just curious.................

[philk]

Dan,

It's me Phil from the workshop on this project.

Tell George not to cut the welds off I made..lol

also Let George know that I warned him about drinking beer while installing the blades. (forgot to tighten them, or was that me).....

I had a blast at your workshop, Learned a lot of things about building wind turbines. And how to make plasma with a microwave..lol

I hope to come buy your place sometime to visit all of you guys again, if you have a work shop at your place again, I would love to help out with it.  

The tail raising up is strange, it would have to lift up at least 2" to get past

the other side of the yaw tube. Must have had a lot of gusty winds slamming the tail around.

As far as the loose blades go, That was George that did that...lol

[frackers]

Nylocks definitely. With the current arrangement of spring washers, if the nuts come loose for whatever reason (vibration, wood shrinkage) then the nut is able to just spin off. With a Nylock then shrinking wood won't allow the nut to move - you might have a bit of a shimmy on the blades but they will stay on!!

Must admit that I built mine with both Nylocks and Loctite Red.

[luckeydog]

Hey Dan B And Tom... thanks for sharing your disaster.  this type of post is my favorite type of post to see.  Everyone here hopefully learns from your experience.

I know I do.

Keep smashing and crashing them so we don't have to.

Luckeydog

.

[Flux]

I prefer to fit a steel plate ( not very thick say 3mm) over the stud area. I think you need very big washers to prevent squashing the wood if you go that way and even with a thinner steel plate ( 1.5mm) I have seen conventional washers indenting the wood.

Nylocks or loctite will prevent things coming off even if it comes loose from compressing the wood ( which the gyroscopic forces will most likely do even if the wood doesn't shrink).

This is a silly issue and can easily be solved, same thing for the tail hinge ( only ever used that construction on 2 tiny machine.

The furling issue is a much more concerning thing to me. I really do wonder how many of these things actually furl rather than rely on stall limiting.

I don't know whether we are going below the critical offset or whether the tails are just far too heavy. I have always got away with 1/" per foot offset but I don't think there is much in hand. One machine  built as a 4ft machine with the intention of modifying the alternator to take 5ft later and I based the offset on this. I got a better result from the alternator modification than expected and went for 5'6" and it doesn't furl. The tail goes over at a fair angle but the power keeps rising. It is on a site with little real wind and hasn't come to harm but I know it wouldn't survive a normal site.

With the inclined hinge restoring force I am convinced that once it does properly furl you will see a drop in power. ( I have to try very hard not to let this drop not to become too great but fiddling with the restoring force). If it doesn't drop then I think you are not furling properly although it may be helping the stall control.

With my methods of loading where I am away from stall the drop in power may be partly due to the fact that once you overcome the seeking force it puts part of the blade in stall and the power drops. There is no real drop in power until the tail gets to about 45 deg and then it seem to defeat the seeking force and power drops.

If the seeking force is much lower with the blades stalled then the transition as it comes into furling may not be so sharp but I can't believe that the power will not drop to some extent.

Flux

[Flux]

I seem to have managed a record number of typos in that lot. Probably most make sense but this is confusing at least to me.

 "I have to try very hard not to let this drop not to become too great but fiddling with the restoring force)."

The but should have said by.

Sorry

Flux

[jmk]

 I used to use lock tight till I stripped the threads on my 5/8 ss all thread. Since then I just put a dab of cheap silicon on the threads when I put it on. I never have had a problem with it yet and when I take it apart it comes apart without problems.

[scoraigwind]

Thanks again for the good pictures that we can all learn from.  Plenty of lessons to be learned all the time from these machines!

It's true as you say that the tail stop was not quite right.  Looking at it closely it doesn't seem to bear squarely onto the yaw bearing.  It's surprising how tricky it is to get the position of that stop right so that it does.  In this case it is bearing onto the side of the tube in such a way that it forces the top of the tail away from the yaw much more than it needs to, and this seems to have caused the tail to lift as you pointed out.  It's like a cam action forcing the tail up on its hinge.  Unless I am not seeing it right and the ail is already lifted way beyond its normal position.

I suspect that the problem would not have occured if the stop had been further out along the tail and bearing squarely onto the yaw.  As a further safety you could design the stop to include a hook shape beyond the yaw that pulls it in rather than pushing hte tail outward.  I am reluctant to add anything to hold the tail down as this tends to make things over-complex and sometimes lets rain into the tail hinge. anyway I feel that a stop that bears both on and beyond the yaw tube would have prevented this from happening and I will look at doing this in future.

I have just been in Nicaragua building another such machine with a few modifications.  The most important mods were to the alternator mount, taking it out from 6 inches to 8 inches offset sideways (my usual offest for a 12 footer is 8 inches) and also pushing the alternator a bit forward toward the wind in its normal position.  I think this helps it to furl properly in the stronger winds.  In the last couple of years I have changed all my larger machines to sit further forward, more like the AWP frame layout.

As for whether the furling system works, I have to say that it does work pretty well, based on thirty years experience, but it does sometimes need some tweaking, and you do need to have a big enough offset.

Moving on to the blade fixings, I rely on the screws rather than bolts although I am sure that the main alternator bolts help to hold the whole thing together.  I never use a steel plate, and I have never seen a hub fail like this that I can remember.  But I will take more than usual care to lock the nut in future!

All good fun

Hugh

[MattM]

Could the tail simply have flexed into the spinning blades and that caused the catastrophic jump out of the tail pivot?  In gusty conditions like that you get unpredictable cross winds that may have put so much force on the tail itself that the fin itself flexed.  The resistance to right itself in the wind due to the inertia force of the spinning rotor and blades may have been what ultimately doomed it in the cross winds.  If that is the case you might need significantly more clearance between the tail and the blades when it furls.

[Flux]

Very extreme things can happen in seriously high winds. At least here the higher the wind the more turbulent it gets. Gusts combined with rapid yaw due to turbulence can have drastic effects.

It looks in this case as though the tail got too close to the blades. It may have been a case of just too close or it may be combined with a fair bit if blade flex making a marginal clearance not enough.

The worst case ought to be clearance to the tower and any blade strike would logically have happened there. In this case it looks as though the tail was closer than the tower because of some failure of the stops. It may indeed have been compounded by some flex in the tail vane but it should never have been that close.

TomW seems to have had a similar failure and from his pictures it seems almost impossible that it could have happened with his stops but it did.

If you aren't using the tail as a back up method of shut down then there is probably no need for it to go to 90 deg. The reduction of power even at 80 deg would probably be adequate so limiting it to about 85deg and having more clearance might be a good idea. With the tail hinge behind the yaw axis and restricting things to 85deg the tail should always stay well behind the tower if the stops work.

I have to admit that I have seen things happen that don't seem possible and unless you happen to see it it takes a very long time to puzzle out what did happen.

I am inclined to think that Hugh's idea of a stop that physically prevents the tail lifting at this extreme limit of travel is a good idea, that way it can't over run the stop even if it tries to lift. If it just contacts a stop there can be very strange vector forces that normally would do nothing but may behave unexpectedly in a violent yaw.

I must confess to have done mods to cure one thing and later found that it has other unforeseen effects so we should at least be grateful when people share their failures and we can learn from it.

Flux

[oztules]

Hugh,

I was playing with the furling of the chainsaw blade mill today in gusty conditions. The tower is still only 9 feet tall... so ground level. Still with the rope on the tail.

I placed a small bucket on the tail rope, and filled it with sand. In the gusts, the power came up to 40A @ 55-57v. eeek

With taking out the sand progressively, it came down to no sand  and so now had max power of 10-15A max @ 53-55v in the gusts... and was furled out enough to completely protect itself.

So some 4kg of sand was the difference between 2kw and 600 watts before furling took hold.

This gives some idea of how iffy it can be to just assume that the tail geometry is close enough... A few pounds here and there or the tail plane a few more inches out or in can make a huge (read terminal) difference.

So your system works, but I'm not sure everyone understands just how critical it is to set up correctly.

(our AWP is going well with it's new blades too. I will be interested to see if they can take the punishment the glass ones did.

..............oztules

[TomW]

Flux;

I added a comment to my Diary with the postmortem forensics on the second:

http://www.fieldlines.com/comments/2008/10/29/163256/68/13#13

That blade hit the back guide rope which the installer [myownself] neglected to secure safely out of the plane of rotation on the prop.

Just to be accurate.

I really wish more folks would cowboy up and share their screw ups. I find you learn more from failure than success, Within limits, of course.

Tom

[pepa]

    Owners fault! This one is for you Tom; I moved the portable test stand to a different area and mounted the mill on it before I installed the sandbags on the base for holding tower down. Went back for the back for bags and oops! I made of three blades for a customer for his porch fans [the wind off a wide marsh creek kept breaking his store bought blades]. I made an extra blade to test as a wind mill that could double as a battery charger in a long power outage and as a fan with dc motor at other times. Works great mounted to his banister rail when used as a windmill... pepa

[TomW]

Pepa;

I am starting to think of them as "Senior Moments" even though I am still [barely] on the under side of sixty! I had a lot going on that day so probably got  in a "horse seeing the barn door after a day of plowing" hurry and cut corners.

I did find the rest of the shards from the prop about 150 yards downwind. Very thin slivers must have rode the wind. I never have found the "Arts Blades" that came off a Zubbly [RIP] conversion 2 years ago in a spectacular failure reinforcing why NOT to use threaded pipe fittings in a yaw setup.

Thanks for sharing.

Tom

[scoraigwind]

Hi oztules

We recently built a 12 footer at blueEnergy in Nicaragua, and put a very light tail on it.  Nice to see it furling at about 500 watts around 7 or 8 m/s windspeed.  Furling does indeed work very well but it does really need 2 things:

sufficient offset and

correct tail moment-or-weight.  

If in doubt offset it plenty and start off with a light tail.

I usually find that wooden blades work best on the AWP but they do have some pretty sturdy GRP blades now.  Not too sure of the shape...  The tip pitch is usually much too coarse for me but they work.

have fun