New Blades for Spirit of Zubbly

[SparWeb]

If you have done it this way, you have not measured the torque at the windmill, so you have not measured the mechanical power directly.

Hi Adriaan,
Yes I have measured torque and RPM, directly and independently.  You should read my page about bench testing my generator to understand what I did.
http://www.sparweb.ca/3_Gen_MoCo/Baldy.html   
Torque was measured directly.  RPM was measured directly.  All was done with calibrated tools.
It's easy to not consider these methods when many members of the forum are hobbyists and builders without formal education in the matter, so they try testing without the engineering. 
Don't hold that against them; they're being creative and still have great ideas.

I have the unfair advantage of well-equipped shops at work that have a wide range of test and measurement equipment at my disposal (sometimes). 
My unfair advantage is compounded because I had the opportunity to run an engine test cell with dynamometer and real time telemetry when in college.
I've been very lucky and it's one of the reasons that I hope that sharing my results with people will help them if they also enjoy studying wind turbines in detail.


On your second point:

You say that the data presented is almost entirely outside the furling range.  This can't be true.

...you could be right.  So I am studying the motion of the tail more carefully today.  Observations on the spot show the tail down at ~250 RPM but the angle of yaw certainly is changing significantly.  So I may have spoken too soon about that.

[SparWeb]

Yes furling is happening in a loop between 250 and 300 RPM. 
And the blades are definitely yawed to the wind direction.
I will keep watching, as you can see in the video below, for specific numbers and range of operation to get it tuned better.
https://youtu.be/VVhAKhYnYpo

Yes, there is vibration visible in the video.

[Adriaan Kragten]

If you want to know if the rotor is really turning out of the wind at high wind speeds, you can mount a wind direction meter on top of the head. This device then measures the difference in between the wind direction and the position of the head if you adjust it such that the given angle is zero if the vane blade of the wind direction meter is in line with the rotor axis.

The inclined hinge main vane safety system is rather complex to describe because the moment around the hinge axis differs from the moment around the tower axis and because the wind speed at the vane blade depends on the position of the vane blade if the vane is normally behind the rotor. The so called self orientating moment of the rotor also has a large influence if the rotor eccentricity is small. So in report KD 431, I could find no mathematical expression for the delta-V curve.

In report KD 377 I have described the pendulum safety system which also uses a weight to counterbalance the rotor moment. This system is easier to describe and it is possible to find a mathematical expression for the delta-V curve. In figure 3 of KD 377, I give the delta-V curve for a certain choice of the parameters. The delta-V curve of the inclined hinge main vane system will have about a similar shape. However, this is only true if the vane bearings have no friction. If the bearings have a lot of friction, you get hysteresis in the vane movement and this can result in an unstable behaviour.

I understand that you haven measured the torque and the rotational speed when you have measured the generator during the bench tests. So then you can determine the supplied mechanical power. If you measure the electrical power, you can determine the efficiency. Most people can measure only the electrical power. But I am not sure if you have also measured the torque when the generator was mounted in the windmill because this requires that the generator housing can rotate or you need a torque shaft in between the rotor and the generator. Both options are rather complex and therefore I think that in the windmill, you have only measured the electrical power and that you used the measured characteristics from the bench tests to find the corresponding mechanical power.

[SparWeb]

But I am not sure if you have also measured the torque when the generator was mounted in the windmill

I am fairly convinced, myself.    It's just a brake dynamometer. 
Once the properties of the generator are measured, it will behave consistently when installed in similar conditions.  I made sure to replicate the conditions of the generator's installation during my brake tests.

[Adriaan Kragten]

But I am not sure if you have also measured the torque when the generator was mounted in the windmill

I am fairly convinced, myself.    It's just a brake dynamometer. 
Once the properties of the generator are measured, it will behave consistently when installed in similar conditions.  I made sure to replicate the conditions of the generator's installation during my brake tests.

This answer makes it clear. You have not measured the torque in the windmill. You have only measured the electrical power and you assume that the conditions in the windmill are the same as during the brake tests. But during the brake tests the generator was loaded with a certain electrical resistance resulting in a certain voltage and a certain current which gives you a certain electrical power. This together with the measured mechanical power results in a certain generator efficiency. You can only use the measured electrical power in the windmill to transfer it to the supplied mechanical power of the rotor if the electrical load is exactly the same as during the brake tests. But this will be difficult to guaranty.

[SparWeb]

Hi Adriaan,
I don't know what to say in balance between correction and politeness.  I've spent 20 minutes trying to decide how to correct you for the Nth time without sounding like a jerk.

My website has been up for more than 10 years, linked to my Fieldlines profile the entire time, regularly referenced in my posts on this thread and many others, and contains all of the details that you are looking for.  And I am certain that you of all people have the technical background to understand all of it.  But you obviously haven't taken the time to read it.  It's very discouraging.  I have read many of your self-published reports.  All I ask is the same courtesy.

http://www.sparweb.ca

[topspeed]

But I am not sure if you have also measured the torque when the generator was mounted in the windmill

I am fairly convinced, myself.    It's just a brake dynamometer. 
Once the properties of the generator are measured, it will behave consistently when installed in similar conditions.  I made sure to replicate the conditions of the generator's installation during my brake tests.

This answer makes it clear. You have not measured the torque in the windmill. You have only measured the electrical power and you assume that the conditions in the windmill are the same as during the brake tests. But during the brake tests the generator was loaded with a certain electrical resistance resulting in a certain voltage and a certain current which gives you a certain electrical power. This together with the measured mechanical power results in a certain generator efficiency. You can only use the measured electrical power in the windmill to transfer it to the supplied mechanical power of the rotor if the electrical load is exactly the same as during the brake tests. But this will be difficult to guaranty.

I was informed to test the torque like this to define the real power of the system.

[Adriaan Kragten]

Hi Adriaan,
I don't know what to say in balance between correction and politeness.  I've spent 20 minutes trying to decide how to correct you for the Nth time without sounding like a jerk.

http://www.sparweb.ca

Or you don't understand what I am saying or I don't understand what you are saying but for the last time I will try to explain my standpoint.

The generator can be tested on a test rig or it can be used in a real wind turbine. If it is used on a test rig, you can measure the mechanical power by measuring the torque and the rotational speed. The torque can be measured the way you did it. You can measure the electrical power by measuring the voltage and the current. If you have measured the mechanical and the electrical power, you can determine the generator efficiency as the electrical power is the mechanical power times the efficiency. The generator efficiency depends on the rotational speed and the kind of electrical load. In KD 78 I have measured a PM-generator for three different load conditions being, constant voltage, constant current and constant resistance. If you compare the efficiency curves for these loads, you see that they are very different. You should test the generator for the load condition for which it is also used in the wind turbine. So if the generator is used in the wind turbine for 24 V battery charging, you should test it on the test rig for a 24 V battery load or for a constant voltage of 26 V which is about the average charging voltage. If you know the Pmech-n curve of the generator for the correct load, you can check if the matching in between the generator and the windmill rotor is acceptable (see KD 35 chapter 8 ).

In the generator is used in the wind turbine, it is very difficult to measure the torque because this means that similar equipment which you have used in the test rig, must also be present in the windmill head. So in the wind turbine, one normally measures only the electrical power and one uses the measured generator efficiency to calculate the mechanical power supplied by the rotor. But this gives only an accurate result if the load conditions in the wind turbine are exactly the same as the load conditions at the test rig.

What I have tried to find in your comments, if you have torque measuring equipment in your wind turbine or not. Up to now I have understood that you have used only torque measuring equipment at the test rig. This means that you need to know the generator efficiency to transform the measured electrical power in the wind turbine into the mechanical power supplied by the rotor. You may not have used the efficiency curves to do this and you may have directly compared the measured electrical curve with the measured mechanical curve but this is another way of using the generator efficiency.

[SparWeb]

Okay...

somehow you can't see something that I can see, or I haven't explained it well enough, or something else is getting in the way; I don't know. 
Here are the pictures copied from my website:

http://www.sparweb.ca/3_Gen_MoCo/Baldy.html
PERFORMANCE TESTING OF THE GENERATOR (June 2010)

Using the lathe is very convenient for me. It's a 5HP power source that can drive my generator at any speed I care to use.

I hauled a whole bunch of batteries with me, and I wired up that board in advance so I wouldn't piss around while at work. The planning paid off because I got in, did all the tests, and packed out in 5 hours. I brought my old Wattmeter with me, and it was indispensable. It's much simpler just reading off that steady needle than trying to decide which of the 3 decimal places to keep on a DMM. Speaking of DMM's, they actually agreed with each other most of the time, a nice surprise. Just to illustrate and have fun, I ran the current (about 40 amps) through a thin piece of wire (last picture) and made it glow nicely!

The reason for all this stuff is so that I can measure both the input torque and RPM. With those figures I have the input power. With the wattmeter on the output side, the analysis gets pretty simple: efficiency is power_out divided by power_in.

All of the tests used 24V batteries. Maybe I could have tried 48V but I can't contemplate a 48V system at home right now, so it would be totally academic. The goal of the tests was to size up the prop, and identify the preferred connection scheme for the generator. I don't know if everyone would agree, but I'm leaning toward Jerry (individually rectifying the phases).

The lathe drives the shaft of the generator.  There is a tachometer on the lathe to give independent confirmation of the speed.

The body of the generator is free to rotate except for the wooden board.  While the generator is forced to turn, the end of the bar rests on a scale which stops the body from turning.  The force at the end of the bar gives the torque.

This is explained in a Youtube video: http://www.youtube.com/watch?v=604dgSMzvu8


Electricity generated by the generator is rectified.
Electrical load is battery charging.
There are some other components but they aren't part of the data collection tests.
A shunt was used for current measurement.  A true-RMS meter measured voltage, and independent of those is an analog wattmeter.

Verbal explanation in a Youtube video http://www.youtube.com/watch?v=baOZut11B2E

So I hope this clarifies.

Here again are the data that was collected from these tests:

http://www.sparweb.ca/3_Gen_MoCo/Baldy.html

What this means is that the data that creates the input power graph is composed of only 3 measurements, taken directly:
RPM (measured by tachometer)
Force (on scale by arm)
Distance (of reaction arm)

Note that none of these measurements are dependent on any electrical operation at the other end.  They are physical measurements taken using the prony brake illustrated earlier.  The tools used for each measurement were tested for accuracy, and in the case of the scale, it was even calibrated.

Further, the data that created the output power graph is composed (again) of direct measurements:
DC voltage
DC current (true RMS)
I admit that the DMM measuring voltage is not calibrated, but at times I swapped them and got the same results on each meter.  And the wattmeter provided confirmation in all cases.

[SparWeb]

In the generator is used in the wind turbine, it is very difficult to measure the torque -

-It certainly is, which is why these thorough bench tests are the best way-

- because this means that similar equipment which you have used in the test rig, must also be present in the windmill head.

-which is exactly what I put into my test configuration.

[Adriaan Kragten]

So my conclusion, which I made earlier, that you have measured only the torque during the tests at the test rig and not during the use of the generator in the wind turbine, is right. But why made this conclusion you so agitated? I had no comment on the way you measured the torque at the test rig. If you have used the same battery load for both tests, you can use the measured Pmech-n and Pel-n curves from the test rig to transform the electrical generator power from the wind turbine into the mechanical power of the rotor.

[bigrockcandymountain]

Do these blades start up better with their deep wide armpits? They look like they would. 

[SparWeb]

They sure do.  It almost never stops turning.  When I can't even feel a breeze at the ground, they'll still be turning.

The previous 8-ft blades could be counted on to be charging if they were turning.  It used to take a noticeable breeze to get them started from a stop.

[SparWeb]

Due for an update.
I have had the tower down for a while making some changes to the tail.  I extended the tail vane to keep the blades facing the wind better. 
It may also delay furling by keeping the tail at the down stop.  This may address the drop in power from 20-25kph as the tail was furling too early.

Just waiting for the wind to die down before raising the tower back up again.

[kitestrings]

I wondered if you wouldn't be tweaking the furling.  Adding two feet to the diameter is a pretty big change, but you must be seeing more power early?

We had fierce winds here Nov 1st that left folks without power for up to four days.  There was another, early winter visit yesterday...less snow than predicted, but very high winds since.

[SparWeb]

Not much wind here since raising the tower back up.
I logged a bit of data for 10-20 kph wind but what I really need is a steady strong blow to find out where the furling point is, now.
The tail vane is 18" farther out the arm.

[kitestrings]

How are the new blades performing?  Any noted changes now that they've been running awhile?  Quieter, noisier?  Furling?  Same great taste; less filling...

[SparWeb]

The blades are great! Thanks for asking.

Now that it's been running almost non-stop for several months...  ... the bearing are shot  :\

Several important changes going from the 8-ft blades to 10-ft blades:

Obviously, much more torque.  The shorted-phases does not maintain a stop in strong winds like it used to.

Previously I had a star-jerry switch which would break the star-point and rectify phases separately.  There is no longer any point in doing this, and the power curve has moved down in RPM so much that flipping to Jerry has not produced a single amp.

When I lower the tower to replace the bearings in the generator, I will also re-wire it for Parallel-Star only.  I won't use any of the spare rectifiers any more, and with 6 down-tower wires, I'll probably pair them up on each phase from generator to rectifier.

I have processed the datalogger output many times now, and the power curve does not have spectacular peaks but it's superb at low wind speeds.  I have never seen a commercially-available WT with any output at 4 mph wind but Zubbly-II does.  So it spins every day and it's only stopped when it's dead quiet around.

Furling keeps it behaving very calmly.  Mostly below 400 RPM, but I've measured some peaks above 500 for less than a second.  In a certain frame of mind, one would want to ease off the furling and let the power peak up a bit more.  But after looking at typical wind speeds in my area it turns out there's not much point.  Raising the peak from 400 to 500 Watts would not matter very often.

I've done some speculative predictions of annual energy output, which comes out around 1000 kWHr per year.  Of course that would depend on my being connected and paid for all those watt-hours, and I'm not.  This is just maintaining a battery charge which often gets topped off during sunny days first anyway.  But man, these batteries sure are happy at float-charge all the time.  But now I'm giving grid-tie much more serious thought.

So all in all, very well.

I gave a tech talk at the Alberta Renewable Energy Association in February: http://www.abrenewableenergy.ca/resources/2020%20Feb%2015%20AREA_Presentation_DIY_Wind_Turbine_2020-Feb.pdf

[bigrockcandymountain]

Wow, power out of 4 mph is awesome.  Hmm, now I'm thinking about blade carving again. 

How long did your bearings last? 

[SparWeb]

Luckily, the last bearing change gave me enough trouble that I posted about it:
https://www.fieldlines.com/index.php/topic,147716.msg1022630.html#msg1022630

That change was done and back up June 2013.  For most of 2019 the turbine was not operating, so let's say 6 years of operation.  I have the replacements on my desk right here, now. 6206+6205.  Just waiting for fair weather to melt the snow drift around the base, then lower the tower again.

[SparWeb]

OK Now I'm mixed up.

When I first converted the Baldor motor, I replaced the bearings then, but didn't give much attention to the type of bearings.  They only lasted 3 years.  Going back through my pictures to answer your question, BRCM, I was sure I had some no-name brand bearings on it to start with.  Then I THOUGHT I replaced them with FAG bearings afterward.

But my pictures tell a different story.  My original assembly has the FAG bearings on.  And it was a FAG bearing I had to cut through 3 years later.  So I must have replaced the FAG's with something else.
What did I replace them with??
Whatever they are, they have lasted 2x as long. 
Now I have 1 more thing to look in to when I get the tower down to change them.

The set of bearings on my desk are also FAG's, so I'm hoping I haven't accidentally bought the duds again.

[wbuffetjr1]

Wow! I would love to start making power at 4mph.

Actually..... I'd like to start making power ANYWHERE less than my current 29mph starting point!! Lol

[mmurray70]

Thats a great job, good to hear they work as well as they look. You should post a 3d cad file

[SparWeb]

Hi Murray,
Actually, there isn't a 3D CAD file.
I could post the AutoCAD file of the 2D templates, if you like.

I started making a 3D cad file in Onshape, but I gave up because something about what I did was very "fragile".  If I changed anything it wouldn't rebuild.  So the model never got finished, and the size and shape at the start of the Onshape idea isn't like what ended up being made.  I think I posted a screen-capture on Fieldlines when I started - but none of it was used.

[mmurray70]

sure, i wouldnt mind taking a look. Were your original blades one solid piece or was it two 2 x 8/10's glued together?

[kitestrings]

The cover picture of your presentation is terrific.  Got'ta love the sheen of those blades.  I also really like your hub plates.  No disrespect, but I never cared for the wooden gussets, or fasteners, that Hugh employed.  It just seemed like the wrong material (to have in tension).

So, the star/jerry switch, was that triggered by speed, or voltage?  Was it just logic in the controller, or is there something that will need to change in the alternator itself?

~ks

[SparWeb]

MMurray70,

Here's a multi-page PDF.
To use it, print it all twice (scale 100%), glue the papers to cardboard, then cut "top" templates out of half of them, and "bottom" templates out of the other half.

In the photo below, you can see both templates superimposed on each other.

 10ft Blade Sections.pdf (209.29 kB - downloaded 69 times.)

[SparWeb]

Kitestrings,
After doing that presentation, the audience kept me for hours afterward with questions.
One pertinent question which got me thinking: "Why not cover the whole blade with that protective tape, not just the leading edges?"

Well, uhhh, yeah...

As for hubs, I wouldn't count on wood lasting very long the way Hugh used it in his workshop books.  But mind you - most of those projects were 4-foot rotors and simplified to the extreme in many ways.

My star/jerry switch was triggered by a human finger, applying a small force on a metallic stick protruding from a small box...  The logic as to when it was turned on and off is still entirely a mystery.   

No seriously I never got around to automating it.  By the time I got my datalogger running with a reliable RPM gauge, in such a way that I could actually measure what was really going on, I discovered that Jerry only offered a better power range only at wind speeds when the tail's furling was starting to operate.  I would have to "let out the tail" if I wanted any improvement from Jerry.  And the 8' blades were already worn out so in my mind I chose to build 10-footers and revisit the Jerry question later.

[mmurray70]

MMurray70,

Here's a multi-page PDF.
To use it, print it all twice (scale 100%), glue the papers to cardboard, then cut "top" templates out of half of them, and "bottom" templates out of the other half.

In the photo below, you can see both templates superimposed on each other.

(Attachment Link)

(Attachment Link)

Thank you! If I ever do tackle a set of blades I will be working with cad and CNC cutting them, but Ill keep this as a reference. Is that a standard airfoil or something you came up with?

Were your originals one solid pc or glued together?

[SparWeb]

NACA6415

[kitestrings]

We're going back to blade tape on the leading edge this round.  I bought a roll of 3", 3M 8672 from Aircraft Spruce on your advice.

Historically, I have a mixed feeling about the stuff.  The frustration is that if you get a small nick or tear in the stuff it can be a real nuisance.  I've seen older turbines where there was some odd whistling or ticking noise traced to that source.

I think the simple answer as to why not use it on the entire surface is that you would be challenged to get it to lie flat, adhere consistently and not develop bubbles.

[SparWeb]

Hair dryer and a roller.

(Not a paint heat gun)

It's very thick.  I've used it since my first turbine and never had a nick yet.  Though on a few aircraft projects it's been nicked, but by handling abuse...

[SparWeb]

6-Month Maintenance Update

Blades look none the worse for wear so far.
Of course, it hasn't been exposed to hail yet.

Items on the agenda: 

• New bearings in the generator
• Remove all components of the tachometer I attempted in 2012 since my current one works better
• Re-wire tower for only 3 power wires (no more need for star/delta/jerry switching)
• Re-wire generator for permanent parallel-star operation only
• Re-varnish blades
• Apply extra erosion protective tape
• Re-grease tail
• Thorough cleaning and inspection of absolutely everything else.

For those who like to keep track, 6 months of operation averaging 200 RPM is about 50 million revolutions.