New Blades for Spirit of Zubbly

[SparWeb]

The Spirit of Zubbly has been flying for 9 years, now.   https://www.fieldlines.com/index.php/topic,147962.0.html
The blades are worn out.  Some hail storms have taken a toll.  I can't show much in pictures taken from the ground, but it will be interesting to compare when I do lower the tower and look at the new blades beside the old ones.

Today I picked up the load of cedar planking that will make the new blades.  1" x 6" planks that will need some planing before lamination.
The previous blades were 8 feet diameter.  After doing some data logging and checking the results, I can tell that it's running below the target TSR.
I will make the new blades 10 feet diameter instead.  The furling works very well, so even if they turn out to be too big, it won't be in much danger.
The spreadsheet from Hugh Piggott tells me that I should give them a chord of at least 12 inches at the root.  Time will tell if I reduce that to 10" or not.
I have some time to mull it over as the wood dries.

[SparWeb]

Yeah, great.  I just noticed in the photo that one of the boards is already split.  It's really hard to find quality cedar planks!

[kitestrings]

So, is this locally sawn lumber or something from away?  It looks like red cedar, which we don't have around here.  We have eastern white cedar.  It's good for many things...mostly fencing, but probably not blades.  I've been messing about with splitting rails for fencing, and finding the grain of the wood in some cases twists more than 90 degrees down the length of a rail (12-14').

Will you just air-dry them?

[SparWeb]

It's western red cedar.  Same as I used for the previous blades.
I only have air-dry - going slowly in a cool garage.

[MattM]

Looking at the planks, it looks like its heavy with cuts from heart wood.  The outer sapwood is structurally the best choices.  The two areas dry at different rates, but your heartwood will often twist and crack at higher rates.  Just food for thought when you pick planks.

[SparWeb]

Making progress.
Worth an update.
I have sorted the planks, arranged them by their grain patterns for lamination, and glued them all together.

I chose to use a System 3 glue (again).  This time it's T88, which has a higher cure temperature than "Cold-Cure".  I thought I'd be fine with the weather, but soon after starting to glue the laminates together the temperature dropped for a week below 10C.  The glue for 1 of the blade blanks took about 48 hours to stop being tacky!


Thoroughly clamped while gluing.


Close-up of the tip.


I figure the knot in this blade blank will attract a lot of attention.  I placed a bit of explanation on the picture, but it deserves more than that...

[SparWeb]

For those who want to do woodworking with cedar, it's becoming a bit of an elite sport.
I hope you can read the prices in the picture below, for CLEAR western red cedar planks.
It's Canadian dollars but any way you slice it, that's too expensive for a project that needs nearly 50 planks.



I had to settle for cedar "fenceboards" from the regular hardware store.  They cost 1/3 as much, so I could justify the time to sort through their stock and still throw away about 25% of the boards I bought.

[Bruce S]

SparWeb;
I have cedar fencing that face the neighbor's yard (looks a ton better than plastic white) , and even I've noticed the prices have gone up.

Will be looking forward to seeing how well the laminate works out.

Cheers
Bruce S

[kitestrings]

"Coffee"...the glue that holds all worth-while projects together.

I see piles of shavings in your future.  Have fun.  Good luck.  ~ks

[DamonHD]

Shavings galore!

Pardon me for being so slow, but I had NO IDEA that that's how you were intending to put the planks together to extract the blade.  Wow!

It almost as if that stuff doesn't grow on trees...

Rgds

Damon

[bigrockcandymountain]

That looks great.  I really like the t88 glue too.  I use it for everything.  Mostly fixing the kids' toys lately.

I hate the cedar quality and price.  I built a small deck last year.  Not one piece is even close to clear. 

I am curious why not just use spruce lumber?

[SparWeb]

Hi everyone, and thanks.

I would not trust that I'm getting spruce even if it's labeled.  It's not common in hardware stores here.  If I find it, usually it's in specialty wood stores like the high-priced one pictured above, and if it's labeled "spruce" it could be any fir tree.  It's also heavier.

I'm surprised by the resilience of the old cedar blades.  They're 10 years old and it was only a hailstorm (the kind that stripped the siding off a side of my house) that spelled their demise.  So I want to stay with cedar for reasons like that.

You could be right - spruce could be cheaper in terms of quality+effort if I value the extra time taken to fuss around with this lumber.

[kitestrings]

Spruce also has one extremely undesirable property IMO.  There is a tendency for knots to loosen as the lumber dries.  Our house originally had spruce siding and those areas that still have spruce have lots of half-round edges where small knots once were, and even holes where a small full round knot has fallen out.  We transitioned to white pine, which is fine for trim and siding, not so good for blades, and thereby largely gotten away from the loose knot issue.

Cedar also has the natural resistance to rot if an exposed portion develops over time.  We only have eastern white cedar locally which is not as strong, or consistent.  I mess with splitting rails for fencing here and the grain often spirals 90-degrees from one end to the other of a 12-14' rail.

We used balsam fir for our blades.  There are some trade offs...but it is local - we cut the logs here - straight-grained, inexpensive and actually quite strong relative to weight.  You have to pick it over and avoid knots, and pitch pockets.  Clear stock is possible as it grows tall with few lower level branches in some areas.  It lacks the resistance to water, decay, so it has to have finish maintained.

[bigrockcandymountain]

You are right that the knots like to fall out.  Both my blade sets have been 2x8 spf lumber.  I found quite nice stuff with vertical grain but it took a lot of picking.  I still had to fill some knot holes with epoxy. 

I have some douglas fir that would be great for blades but we bought it for trim and i am too lazy so far to laminate it.  It is super tight grained and clear.  I think it would  be better than spruce. 

Cedar smells so nice when you cut it that it is almost worth it just for that.

[SparWeb]

Got the front face of 2 done and sanded to profile so far.

FYI the blanks started out at 36 pounds each.  Curious how little will be left when I am done

[Adriaan Kragten]

Dear Spar Web. In your first post you say: It's running below the target TSR. So I will make the new blades 10 feet in stead of 8 feet". I am afraid that increase of the rotor diameter with a factor 5/4 is too much to get an optimal matching in between rotor and generator. Matching is explained in chapter 8 of my public report KD 35.

If the rotor is running below the target TSR it means that the optimum cubic line of the rotor is lying below the Pmech-n curve of the generator for moderate wind speeds. The formula for the optimum cubic line is given as formula 8.1 in KD 35. In this formula you can see that the rotor power P for a certain Cp, a certain rotational speed n and a certain TSR lambda increases by R^5. So if the rotor diameter is increased by a factor 5/4, P for a certain rotational speed will increase by a factor (5/4)^5 = 3.052. I expect that with this strong increase of P, the optimum cubic line of the rotor will now lie far above the Pmech-n curve of the generator. This means that now the rotor will run above the target TSR. For the correct increase of the rotor diameter you need to know how much the optimum cubic line is lying below the Pmech-n curve of the generator for the 8 feet rotor.

I don't understand why American people still use non metric values like inch, foot, mile, pound and hour if they talk about wind turbines. If you use the metric system, all formulas about kinetic energy, mechanical energy, potential energy and electrical energy are very simple and one kind of energy can be transfered easily into another kind of energy (see public report KD 378).

[Bruce S]

I don't understand why American people still use non metric values like inch, foot, mile, pound and hour if they talk about wind turbines. If you use the metric system, all formulas about kinetic energy, mechanical energy, potential energy and electrical energy are very simple and one kind of energy can be transfered easily into another kind of energy (see public report KD 378).

SpareWeb
I'm confused , your flag shows Canadian  .

Happy Monday!!

Bruce S 

[topspeed]

The Spirit of Zubbly has been flying for 9 years, now.   https://www.fieldlines.com/index.php/topic,147962.0.html
The blades are worn out.  Some hail storms have taken a toll.  I can't show much in pictures taken from the ground, but it will be interesting to compare when I do lower the tower and look at the new blades beside the old ones.

Today I picked up the load of cedar planking that will make the new blades.  1" x 6" planks that will need some planing before lamination.
The previous blades were 8 feet diameter.  After doing some data logging and checking the results, I can tell that it's running below the target TSR.
I will make the new blades 10 feet diameter instead.  The furling works very well, so even if they turn out to be too big, it won't be in much danger.
The spreadsheet from Hugh Piggott tells me that I should give them a chord of at least 12 inches at the root.  Time will tell if I reduce that to 10" or not.
I have some time to mull it over as the wood dries.

(Attachment Link)

How big were the hails that caused the damage ?

[bigrockcandymountain]

I would love to see pictures of the blade carving. 

Wait, are hours metric or imperial

[SparWeb]

Hi everyone!
Answers in order then-

Hello Adriaan,
I guess I could furnish more detail about the blade-to-generator matching process.  My approach is very different from yours.
My previous 2.44 meter diameter blades were apparently under-sized for the turbine, and for that it really didn't matter if they had been cut to the correct TSR or not (which they weren't).  When I first cut them, I was still playing around with the math learning to understand it.  Today I have all the resources I need to work out exactly what's needed, and I wouldn't mind sharing it, if you like. 

I use software known as "Mathcad" which permits analysis of engineering problems with any unit system I want. You can be confident that I use metric for my private aerodynamic calculations, but for my mostly north-american audience I usually report measurements in US customary units when I post to Fieldlines.  This is probably not widely known, and when I explain it, sometimes it falls on deaf ears, but ever since Canada chose to use metric, but the US chose NOT to use metric, Canada has been in a units "purgatory" where we have to be aware of both and express ourselves one way for one person and a different way for another.  I'm used to it, although I admit the US customary stuff still comes more easily.

Anyway, back to turbines:
I had measured, directly, the power input and output curves of my generator by driving it using a lathe.  I am confident I've mentioned this to you before, and I documented some of this on Fieldlines, if you want to see my test apparatus and methods.  The data allows me to compare many different blade profile and power output curves, to see which ones match the required input of this generator.  I can even select different electrical connection schemes of the generator to improve the match, rather than choose different blades, if I want.

The essential conclusion is that a 2.4 meter rotor is too small, and this was borne out by the data I collected with my datalogger last year.  Yes it took me several years to figure out how to built a suitable datalogger, but once it was done, I was able to observe the turbine running well below its design TSR, and nowhere near the most effective speed for the generator.  Referring to the generator's power curves, you can see that an output of 600 Watts or even 1kW is possible and quite safe to do from time to time.  However, with the 2.4m diameter rotor it rarely made more than 500W.  I was simply too conservative on my first set of blades.

There are other reasons... I have replaced the generator several times, too...  it's a long story and also on my website if you choose to explore it.

[SparWeb]

Bruce, don't judge until you've walked 1.6 km in another man's shoes.
BRCM, I'm just used to it.  I work in metric 10 hours a day, the other half in imperial.

[SparWeb]

The hail storm that brought the end of these blades so much closer happened in 2016.  Hard enough to take the siding of my house, and the houses of all my neighbours.  Everybody was visited by the insurance adjusters for a few weeks.  My house needed the siding and roof.  Pictures below.

It was almost 3 years ago exactly - chipping and splitting most of the trailing edges and leaving a number of dimples.  The leading edges survived and they did not blow up in the windstorm that was happening at the same time.  I was preparing to take it down after the storm but life was extremely busy at the time, I wasn't even at home very often, and a few more windy days proved that the blades still had enough integrity to stay up.

But it was clear that they had suffered a great deal.  I might have done this carving last year but I had some other big-ticket things to take care of. 

Pictures from the storm below.  They were golf-ball sized but the destructive part was the size and tremendous speed they hit.

...sorry no pictures yet - can't get this to post with them attached...

[SparWeb]

Photo during the storm

[SparWeb]

Hard enough to peel the potatoes in the field

[SparWeb]

Damage to the blades.  It's been running like this for 3 years.  Fingers crossed!

[SparWeb]

Right:
Actual blade carving pictures!

Working with a short draw-knife:


And with a long one:


Complicated geometry.  Need to lop off a bit triangular shape where the curved sweep of the blade airfoil blends into the thick solid grip at the hub.


I refer to this area as "the armpit".  Sorry.
And, of course, as I trim and file it smooth, that means I am "shaving the armpit".  double sorry, couldn't resist.

[SparWeb]

Getting the profile right is difficult with the tools I have, because the airfoil (NACA 6415) has a slight concave curvature on its bottom face.
I have a few small spoke-shaves that help with this.

The templates were drawn in CAD, printed on paper at scale 1:1, then glued to sheets of cardboard so that they stand up.

At this point, I have 2 blades "close" to right on the template curves for 2 blades.  Which puts me about 1/4 of the way through the carving process.

[Adriaan Kragten]

SparWeb, if I look at your graph, I see that you compare the Pmech-n curves of the generator for a certain load with the curve of maximum Cp for the rotor. This curve, I call the optimum cubic line. So you are doing the same as me when you determine the matching in between rotor and generator.

What I find strange is the big distance in between the P-n curves of the rotor for high wind speeds. It seems to me that you have determined these curves for a rotor perpendicular to the wind. But your wind turbine has a safety system which turns the rotor out of the wind at high wind speeds and above a certain yaw angle, this results in decrease of the distance in between the P-n curves of the rotor. For the safety system used in my VIRYA windturbines, the rotor turns out of the wind at high wind speeds such that the P-n curves of the rotor for high wind speeds are even about the same.

For the hinged side vane safety system used in my VIRYA wind turbines, it is rather easy to determine the graph which gives the yaw angle delta as a function of the wind speed V (see KD 213). So the effect of this delta-V curve can be incorporated in the P-n curves of a yawing rotor for high wind speeds if you use the formulas for the rotational speed and the power as given in chapter 7 of KD 35. For the inclined hinge main vane safety system as used in the design of Hugh Piggott, determination of the delta-V curve is very difficult. I have tried to derive the formulas in public report KD 431. I found the moment equations around the tower axis and the hinge axis but I could not find a formula for the delta-V curve.

[bigrockcandymountain]

Cool i wondered if they would be a flat bottom airfoil or cambered.  I had ideas of using the goe 222 but i was in a hurry carving, so they are flat bottom. 

That is very ambitious carving a camber.  I would try an angle grinder with a flap disc sanding pad.  You obviously got it figured with a spokeshave which is way more peaceful. 

I was through your area just after that hail storm.  It was the worst damage i have ever seen first hand.  I'm impressed that your turbine made it through. 

[SparWeb]

I would try an angle grinder with a flap disc sanding pad.  You obviously got it figured with a spokeshave which is way more peaceful. 

No - I need the quiet.  On many levels.

I was through your area just after that hail storm.  It was the worst damage i have ever seen first hand.  I'm impressed that your turbine made it through. 

Yea I didn't believe it was going to work, and had it shut down for a while.  Then I tried running it a little, tested for vibrations coming through the tower.  Nothing had changed.

FYI, there is a simple instrument that you can use to estimate the intensity of vibrations: lean you head against the tower!  So much noise it transmitted to your skull.  Then think of a song with the same tempo and key as the vibrations you hear.  It was still playing Kim Mitchell's "Soda" so everything was fine.  If it starts playing Alice in Chains, then I've got trouble.

[SparWeb]

...the big distance in between the P-n curves of the rotor for high wind speeds. It seems to me that you have determined these curves for a rotor perpendicular to the wind. But your wind turbine has a safety system which turns the rotor out of the wind at high wind speeds and above a certain yaw angle, this results in decrease of the distance in between the P-n curves of the rotor.

Hi Adriaan,

Thanks for the comparison.  The higher power curve at higher wind speeds is, of course, due to the wind speed's cubic power-law driving the turbine.  In any yawing, the incoming power is reduced, and faced with the same generator power curve, the blades operate at higher angle of attack until they stall.  Since this is a safe result, and exactly the intent of furling after all, then I don't see much need to analyze it.

Some analysis of the effectiveness of the furling tail, such as its activation speed and angle of its travel is useful, though.  With a larger rotor my system is receiving higher thrust and the rotor will also have a higher inertia.  Both of those factors will affect the furling behaviour.  In some cases they may cancel each other, in other cases, they may reinforce. 

On the one hand, increased thrust will only happen if the rotor is more highly loaded.  Since my generator power curve remains close to the rotor's power curve, that will happen.  I can expect that force to drive the furling earlier, or at a lower wind speed.
On the other hand, the higher inertia of the rotor does not allow it to yaw as quickly.  By yawing more slowly, it will be less sensitive to gusts.  Currently, I feel that the WT is too sensitive to gusts, so I believe this would be a positive change.

For now, I do not see a safety-related reason to change the furling or the balancing mass on the tail.  To optimize performance, and to realize the full benefit of the increased rotor size, I may choose to add weight to the tail, thus delaying the furling to a higher wind speed.  I would like to take datalogger readings of the performance before I do that, however.

[kitestrings]

Looking good Spar.

I found the drum end of my belt sander works pretty well on those "pits" near the root of the blade.

Looking forward to seeing the final product.  ~ks

[SparWeb]

Due for an update!

I got the front sides down to the rough profile a couple of weeks ago.  This is the "flat" bottom of the airfoil, which isn't actually flat in this case.



Since then I've been working more slowly on the rear sides; the curved "top" of the airfoil.



The paper templates fit together, so putting the top and bottom templates together I can tell how close I am to the final profile.  I got all of the profiles "really close" before going back over them all again and doing a pass to "very close".



At this point, the airfoils only need sanding.  Then to trim down the hub grip ends.

My band-saw is struggling to cut through the thickness.  Slow and steady...