PVC turbine rebuild - looking for recommendations

[Ontheronix]

Hi all!
I will try to repair and improve a windturbine from a local, creative bar in Antwerp, Belgium:

The windturbine:



One of the PVC-blades is broken, probably from freewheeling. The alternator seems to be intact: each phase measures 1,5 ohm and the bearings/tolerances are ok. The original creator also had an enclosure built around it:






I tried to match it to the Hugh Piggot design, but it seems to be a bit smaller:
Neo: magnets are 66% smaller, but the rotor has an extra stack
Rotor: 20cm, whereas Piggot's rotor is 23 cm (1,2m single rotor windturbine) or 26cm (1,8m dual rotor windturbine)
Wire diameter: unknown

Due to the ineffeciency of the PVC-blades, the stator probably never burned out. (It has no furling or electronic short-circuit/-brake whatsoever). I'll upload a picture of the old PVC-blades but the root had a curve of 90° which was acting as brake. I don't have a crazy amount of time so I will make PVC-blades again instead of wooden ones. I will use the PVC-blade-calculator from Wind&Wet: http://www.windandwet.com/windturbine/tube_blade/index.php
The bar is located in the outskirt of the city, at the harbour, so it has some good wind coming in. Since the tower is only ~5m high, I estimate the average windspeed at 3,5-4m/s. The map states 4,1m/s:


I'm looking for some recommendations for the length of the new blades. Originally they had a length of 58cm. But Hugh Piggot uses the same volume of magnets for his 1,8m turbine, but the diameter of my alternator is smaller (see above).
Also I am wondering how bad an alternator encasing is. Since the rotor disks are rusty it will be hard to give it a decent paintjob to make it weather resistant (we have 800mm/year rain here ) Other ideas and general recommendations also more than welcome!
The most important thing for this project is showcasing that windturbines can be made by everyone. Electricityproduction is welcome but not top priority.

[bigrockcandymountain]

If electricity production isnt important, carve some rough wood ones.  I use a chainsaw to start with and a power planer.  You could use almost any tool to make straight, not tapered, not twisted blades.  Table saw, circular saw etc.  Maybe others on here have a good plan for pvc blades, but most i have seen aren't durable. 

In that location, a blade flying off could kill someone.  Just a thought.

[Bruce S]

While looking at the pics of the NEOs and the steel plates, I'm impressed that the magnets aren't corroded.
IF you want to go the extra mile (Km) you could carefully disassemble that, use some fine grit sandpaper and get it polished up along with getting it painted.

Give those magnets a good close look at too, make sure they're not showing signs of having the Nickel coating coming off.

In looking at the other two blades, you could make a template from those and put up 3 new ones, however, before you put it back up. Paint the PVC, most PVC like those are not UV protected and most likely the reason the 3rd one broke off.

Those also look to have been cut short on purpose. This would've been to keep it's RPMs lower and to keep them from smacking that light pole.
It was put up there probably as a conversation piece, and maybe to light a LED or two.

BTW!!: When did this originally go up? Antwerp is a beautiful place!! Haven't gotten to go visit in a good 8 years.

Hope this helps

Cheers
Bruce S

 

[Ontheronix]

If electricity production isnt important, carve some rough wood ones.  I use a chainsaw to start with and a power planer.  You could use almost any tool to make straight, not tapered, not twisted blades.  Table saw, circular saw etc.  Maybe others on here have a good plan for pvc blades, but most i have seen aren't durable. 

In that location, a blade flying off could kill someone.  Just a thought.

Safety before everything! How big do you think the difference in performance is? Let's say I make a small version of this one:

I plan to make an electronic brake in the near future (an Arduino sensing overvoltage -> make a transistor short the phases).
PVC blades have a bad name in terms of performance, but some counterexamples are also there. Maybe it's an idea to make some PVC-blades now, and try to CNC some good wooden ones when I have access to a CNC-machine (in 2 months)?
It would be nice of the windturbine produces something usable though. It doesn't have to be top notch in terms of efficiency but it would be nice if it could charge a carbattery a day.

While looking at the pics of the NEOs and the steel plates, I'm impressed that the magnets aren't corroded.

The rotor and stator were encased in a box, so no rain got to them. The steel discs were probably rusty to begin with

IF you want to go the extra mile (Km) you could carefully disassemble that, use some fine grit sandpaper and get it polished up along with getting it painted.

So rebuilding the enclosure is really not an option in terms of cooling?

BTW!!: When did this originally go up? Antwerp is a beautiful place!! Haven't gotten to go visit in a good 8 years.

Let me know when you're in the city, we should go for drinks! Surprisingly it has been running for 2 years, and has been down for more than a year now.
The original builder didn't know much about windturbines and merely trew some stuff together as an experiment.

[SparWeb]

It looks easy to access for changes after you set it up the first time.
You could try 1.5m diameter rotor, and if it seems to run too fast, cut the tips off.

I would recommend you make the plywood blades shown above, rather than PVC pipe blades.  Just make them a bit shorter than shown in that diagram.  You can also round the tips a little to reduce the noise.

[Ontheronix]

Allright, will do!  It's easy to access indeed, and you can't fall from very high.
What is your reason to choose straight plywood blades over PVC blades?

[mbouwer]

Or just shape polyester blades?

[Bruce S]

You could go with cooling through the box, but ,, with the air being moist , the equipment could very well continue to rust.
The magnets could also start rusting , this would certain be a bad thing  .

attacking the rust now and getting it stopped will ensure a much longer life ( Possibly even long enough for me to make it back over  )

PVC breaks down over time, it gets brittle , start to crack and there goes one of the blades.
Carving a set of blades from solid wood blanks would be the best way to go, but , that takes time and a certain finesse.
Making plywood blades can be just as inexpensive and PVC since these can even be made from scraps.

 

[Adriaan Kragten]

On my website www.kdwindturbines.nl you find two folders at the menu VIRYA-folders in which in totally eleven small wind turbines are described. Detailed technical drawings of the rotors are available for free in the manual of the wind turbine or in the design report of the rotor. An overview of the specifications of all eleven wind turbines is given in the folder "Extended specification". The smallest rotor has a diameter of 0.98 m. The biggest rotor has a diameter of 2.22 m. May be you can use the rotor of one on these wind turbines. However, realise that these rotors are meant for a wind turbine equipped with a safety system which turns the rotor out of the wind at high wind speeds. 

I have bad experiencies with thin tapered PVC blades because of the sensibility for flutter at high tip speeds. Some information about these experiencies and about tapered stainless steel blades is given in public report KD 616.

[Adriaan Kragten]

In my previous post, I have said that PVC blades are sensible to flutter which is mainly caused by the low modulus of elasticity of PVC resulting in a low torsion stiffness. But flutter can be a problem for all long and thin blades and therefore I will explain what flutter is. I have observed flutter in the wind tunnel for some of the rotors as described in report KD 616.

Every blade has a certain natural frequency for bending and a certain natural frequency for torsion. If the blade isn't rotating, the natutural frequency for torsion is much higher than for bending. The natural frequencies depend on the torsion and the bending stiffness and increase if the stiffness is higher. If the blade is rotating, the centrifugal force in the blade has a tendency to stretch the blade and this stretching results in increase of the bending stiffness and so in increase of the natural frequency for bending. Rotation of the blade has no influence on the natural frequency for torsion.

For long thin blades, the torsion stiffness is rather low and there will be a rotational speed for which the natural frequency for bending becomes equal to the natural frequency for torsion. Flutter will start at this rotational speed and the following will happen.

There are always vibrations of the blade due to turbulence in the wind. Assume that this vibration results in decrease of the blade angle beta, so in increase of the angle of attack alpha. Increase of alfa results in increase of the lift L. Increase of L results in bending backwards of the blade. The resulting lift on an airfoil exerts about at the quart chord point so before the neutral line for torsion which lies half way the chord for a cambered plate airfoil. So increase of the lift also results in increase of alfa. So this results in more increase of L and in more bending backwards of the blade. The airfoil will stall at a certain angle alpha and L will decrease at higher angles. So then the blade starts moving forwards.

If the natural frequency for bending and torsion are equal, this bending and torsion of the blade are in phase to each other and the movements are therefore strengthening each other. If flutter starts, the rotor becomes very noisy and if you look at the rotor from besides, you see that the rotor disk becomes very thick at the blade tip.

The airfoil drag becomes rather high when the airfoil is stalling and this increased drag results in decrease of the rotational speed. This decrease of the rotational speed results in decrease of the bending stiffness and if the rotational speed has decreased enough, the whole flutter situation suddenly stops. But now the rotor will accelerate and flutter will start again at a somewhat higher rotational speed. This starting and stopping of flutter can be observed very will for a rotor in the wind tunnel because a wind tunnel gives a very constant wind speed.

The scale laws for flutter say that if all dimensions of a rotor are scaled up with a factor i and if the blade material is the same, flutter will start at the same tip speed. So the rotational speed for which flutter starts will be a factor i lower if the scale factor is i.

[Bruce S]

Adriaan;
Very nice! a detailed explanation of flutter has long been needed.
Especially with PVC blades !

Cheers
Bruce S

[bigrockcandymountain]

That is a very good explaination of flutter.  I will keep that in mind also. 

I think rebuilding this turbine with those plans for plywood blades is a great idea.  I would use solid wood though, instead of plywood, for a couple reasons.

The blades should be stronger.  Plywood has half the plys running the wrong direction so is only half as strong as a straight grained wood blade.

Plywood will have a lot of end grain exposed that will want to suck up moisture.  It will also make carving, sanding etc more difficult.

Don't let me discourage you.  Whatever you decide to do will be great.  Keep us posted. 

[Adriaan Kragten]

I agree with some of the arguments against the use of plywood for use as material for rotor blades. As only half of the plywood layers have the good direction to take the stress which is caused by the centrifugal force and the bending moment in the blade, plywood blades are less strong and less stiff than massive wooden blades made from wood of the same density and with the same geometry. However, plywood can be used for blades which are relatively short with respect to the used blade chord and thickness and for which the stress is therefore not critical.

I have used plywood for the 2-bladed VIRYA-1.8W rotor described in public report KD 664. The advantage of plywood is that it is standard available all over the world and that good quality hard wood is difficult to obtain in many countries like India. Machining the required airfoil in plywood might be difficult with a hand plane but I think that if an electric planing machine is used, it is rather easy, especially if the blades have a constant chord and no twist. Plywood should be protected very well by at least two layers of epoxy and two layers of aluminium paint.

The technical drawing of the VIRYA-1.8W blade is given on drawing 1801-02 which can be found at the end of report KD 664. The blade has a thickness of 18 mm, a chord of 150 mm and a length of 800 mm. So 24 blades can be made from a standard sheet size 1.22 * 2.44 m (4' * 8') with only a little waste material. The two blades are connected to each other by a stainless steel strip size 3 * 90 * 500 mm. Both ends of this strip are twisted 7.5° to give the blade the correct blade angle. The strip is rather elastic and this compensates most of the vibrations in the rotor shaft caused by the gyroscopic moment of a 2-bladed rotor.

[Ontheronix]

I have also experienced flutter with flat aluminium blades in a mini-mill of 70cm diameter, with flat aluminium blades of .5mm. It behaves exactly as Adriaan describes; at higher speeds the blades start vibrating, the rotor becomes a thick disk seen from the side and fails to increase further in speed. And of course there is the noise :p

Is it an idea to carve a GOE222 profile? I have read these have no twist and perform well in lower wind speeds, but less in higher winds. The lack of twist makes it easier to shape. I have quite some solid wood available also, so no need to use plywood for me

Meanwhile I have carved some PVC-blades with the template of Wetandwind to be able to test the alternator and electronics. They need to be replaced by better ones in any case. Certainly because one blade is almost 10% lighter than the other ones, although tolerances are maximum 1mm, so I don't get why one is so much lighter. They weigh 327g, 343g and 347g. I must say the pipe has been laying in the sun for mor than a year, so the only thing I can think off is the side that catched the sun has dried out in some way. (No worries - I will use these only in low winds / controlled environment for a few days).

[Bruce S]

Ontheronix;
The experience you've seen with Aluminum blades is pretty much why, most here shy away from it, the stresses put on the Al will cause early failure due to flexing stresses put on the metal.

I'm curious to see the PVC blades you've already cut.

Good to know you have access to solid wood. Sparweb has a nice post going about laminating solid wood and carving a nice set of blades.

Cheers
Bruce S   

[Ontheronix]

I only made a movie of my test-setup today, I'll make some decent pictures of the PVC test-blades tomorrow. The whole thing is tied on a shopping cart for now (The curvature of the blades is because of the cameralens)

We had some strongish winds of 4,7m/s average today (@10m height). The location is next to a dock, so the wind should be quite laminar. Performance was not to be blown away though:
1A @ 14V (RPM guess with the naked eye: 60RPM)
0,5A @ 28V (RPM guess with the naked eye: 120RPM)

Even unconnected it barely reached 30V, although to my latest information it should be a 24V stator.
One of the bearings makes some noise so I'll take a look at it tomorrow. Let's say that gives me an extra 5W, and wooden blades another 10W? That comes closer to the 30W it should make in those winds (derived from Hugh's 1,2m single rotor turbine). Or maybe I am being to bullish here, and was the wind at ground level lower then I thought.

Meanwhile I dream of making perfect wooden blades at night  I'll check Sparwebs post tomorrow!

[Ontheronix]

Folks! I owe you a follow up! I had a very hard deadline because I was moving out of my apartment, so I didn't get alot of time to post.
I sticked with the PVC blades, which are driving a 40W 24V led string as a dumpload at this very moment!

[SparWeb]

Excellent!
The light show is a nice touch.