generators: ideal target rpm for target voltage

[brandnewb]

I am constructing a generator for a VAWT.

I am explaining more about it in the following thread.

https://www.fieldlines.com/index.php/topic,150595.0.html

It's going to be coil winding time soon and I have no clear idea yet on what I should be going for in terms of voltage production at a given rpm.

The generator and it's drive (the VAWT blades)
of which more can be found here
https://www.fieldlines.com/index.php/topic,150529.0.html
should never rotate faster than 120 rpm to produce the intended 58.4 open circuit voltage I have set as the goal. The turbine will be operating in an urban environment and if the blades would detach and kill someone then my life and that of many others will be ruined.

I could try to wind coils that reach that target at 60 rpm's or I could wind coils that reach that target at 120 rpm's.
Given my safety concerns I am leaning towards 60 rpm's if that is even possible with the generator I am building.

Any thoughts on this would be greatly appreciated.
My

[electrondady1]

i build vertical mills.  i try to produce 12 volts @ 60 rpm. its a use useful voltage at about 1/3 of their  top rpm

[brandnewb]

i build vertical mills.  i try to produce 12 volts @ 60 rpm. its a use useful voltage at about 1/3 of their  top rpm

May I please ask you to elaborate much more? 12V why? 1/3rd how come? etc etc.

[MagnetJuice]

I can estimate the number of turns per coil, but I need more information.

1 – Is this going to be 3-phase or single phase?
2 – What is the size of each magnet?
3 – What is the strength of the magnets; N42, N45, N50?

Once I have that information, I can design for 60 or 120 RPM.

Ed

[brandnewb]

I can estimate the number of turns per coil, but I need more information.

1 – Is this going to be 3-phase or single phase?
2 – What is the size of each magnet?
3 – What is the strength of the magnets; N42, N45, N50?

Once I have that information, I can design for 60 or 120 RPM.

Ed

wow that would be amazing!
The magnets are N45 class (12 KG holding capacity) with dimensions 60x10x5 mm.

How would you go about getting that number if I may be so blund to ask as I am really curious.

EDIT: I forgot to mention that I am I am aiming for 3 phase per coil disk

[MagnetJuice]

With your present magnet arrangement, for 3-phases, you will need 72 coils with 250 turns each to output 58 vdc into a 48 volt battery bank at 60 RPM. As the RPM increases, the battery will clamp the voltage at about 58 volts, and the amperage will keep increasing into the battery.

That will give you a maximum output of about 455 watts. That means that the wire gauge should be able to handle about 9 amps of current. Therefore, wire thickness will be 21 Ga. (.7mm)

I think that it will impossible to fit 72 coils with 250 turns each in the space that you have now.

Assuming that you manage to squeeze those coils in that space, there is bound to be some voltage cancellation because of the way the magnets are arranged. Therefore, much less than the maximum of 455 watts will be produced.

96 magnets of that size have the potential to produce 1,250 watts maximum. To get that much power, the coils would have to be placed between 2 magnets. That way the coil will be located where the maximum flux is.

With the way that the magnets are arranged in your drawing, less than 40% of the magnets flux will be reaching the coils.

Ed

[brandnewb]

this is both good and bad news ;(

It tells me what to aim for and also that I am most likely wasting time.

But I am not shy of wasting time. Can one please tell me where I can find generator simulation software? I'd like to see what theoretically happens if we go half and half between traditional and repelling.

Rotating the magnets 45 degrees. Both in a traditional sense and in a repelling sense.

The reason for me not letting go yet of an unconventional arrangement is because since I am 3d printing i'll need some more material than 5 mm to get some sturdiness.
EDIT: agreed, one could argue to simply get thicker magnets. But I would like to see where this all leads me with the magnets I have got at the moment.

[brandnewb]

I tried
www.femm.foster-miller.net/index.html
but that seems defunct by now.

[dnix71]

Google's Wayback has the site and software archived. I downloaded and installed FEMM from here.
https://web.archive.org/web/20040402184002/http://femm.foster-miller.net/download.htm

[Mary B]

Someone came up with a spreadsheet way back when that got you in the ballpark. Called number of turns, I *think* I have a copy of it still on this hard drive. A quick search didn't find it but I KNOW I saw that file recently... maybe on an old drive...

Here is a discussion in it https://www.fieldlines.com/index.php?topic=139595.0 but no DL available on that one a deep search of the site may turn it up...

[MagnetJuice]

The FEMM has a new website and it has newer versions of the program in 32bit and 64 bit.

You can find it here:

https://www.femm.info/wiki/HomePage

brandnewb, I hope that you download it and use it. Maybe you can find a magnet arraignment that is a breakthrough.

Ed

[joestue]

i am of the strong opinion that the radial to axial transition topology that is being discussed here is a step backwards for low speed machines.

for very high speed machines i think it may be of some benefit if you use a laminated trapezoid metal pole piece surrounded by 3 or 4 magnets. the reason why is because the outer ring could be made of carbon fiber and restrain the magnets at a very high rpm. the pole piece being made from a laminated stock would suffer substantially lower eddy current compared to a single axially oriented magnet which is basically a short circuit (unless you break the magnet into smaller pieces). but for low speed machines the eddy current losses in the magnets are negligible.

what i think would be substantially beneficial would be to investigate using the 30% iron powered filled PLA as a core to wind the coils on. this should boost the output of the machine by something on the order of 30%.. and/or you would need 30% less magnet.

[brandnewb]

i am of the strong opinion that the radial to axial transition topology that is being discussed here is a step backwards for low speed machines.

for very high speed machines i think it may be of some benefit if you use a laminated trapezoid metal pole piece surrounded by 3 or 4 magnets. the reason why is because the outer ring could be made of carbon fiber and restrain the magnets at a very high rpm. the pole piece being made from a laminated stock would suffer substantially lower eddy current compared to a single axially oriented magnet which is basically a short circuit (unless you break the magnet into smaller pieces). but for low speed machines the eddy current losses in the magnets are negligible.

what i think would be substantially beneficial would be to investigate using the 30% iron powered filled PLA as a core to wind the coils on. this should boost the output of the machine by something on the order of 30%.. and/or you would need 30% less magnet.

It could well be that this is a dead end indeed. It all started for me trying to reduce the diameter of an earlier generator which had a diameter of a bit over 80 cm. So I tried to see what repelling could do and found flux values that seem to suggest it's worth a try even though I could arrange traditional with a similar spacing but then loose material to hang magnets in.

Regarding coil cores. I am already experimenting with resin + iron powder getting the iron powder content near 95%. Let's see where this all leads.

[brandnewb]

for the coil test I'd like to create a single coil with 10 winds and spin the magnets at 60 rpm and measure what open circuit voltage I am getting.

I would like to make maximal use of the available flux, which my findings seem to suggest are more than traditionally arranged magnets at close distances.

So I'd like to try the following coil


this extends the coils quite a bit in a northern and southern direction in the respective flux directions. I don't know if the coils on the inner edge really do anything as they do not extend in a northern / southern direction as I don't have space there.

My question, does this coil path make sense? Basically worth a try? Or perhaps I misunderstood some of the finer points behind coil physics.

[MattM]

Your stronger field is probably shifted 30 degrees off the vertical.

We do know that arrangement is 2/3rds of a Halbach array.  It makes me think you could have two stators with magnets turned in a more conventional way, but shifted so that opposing magnets are never directly across.  As the main rotor turns your psuedo-Halbach arrays should rapidly flip magnetic  directions.  The upside is far fewer magnets are needed in a stator.  The downside is, about the only coil that would work in the tight space is a wave winding between your magnets on the stators.  If your numbers of magnets on a stator matched your central array it will probably cog like mad.

[brandnewb]

I've had some more time to let this sink in in the meantime.

Does this mean that if I would arrange the magnets traditionally, which can be done in the same diameter I could theoretically get 1250 W as opposed to 455W using a repelled arrangement? If so it would be a waste to throw away those Watts since there is not diameter benefit.
It would mean however I am not sure I can get the magnet disk strong enough to stay stable while under load.

Or does it mean that when I use 48 magnets on each side as rotors of a single coil stator disk, arranged to attract could theoretically get 1250 W. In that case there will be no problem getting the magnets disks to be strong.

It would mean however needing to increase the rpm right? since the field reversal frequency would be going down by 50%. But then again that could be offset by making bigger coils as now there is more space for that.

Either way I will still collect data to see what will happen on the track I am now (96 repelling 6mm spacing at inner edge) and share that whether bad or interesting and perhaps try something else after that.

With your present magnet arrangement, for 3-phases, you will need 72 coils with 250 turns each to output 58 vdc into a 48 volt battery bank at 60 RPM. As the RPM increases, the battery will clamp the voltage at about 58 volts, and the amperage will keep increasing into the battery.

That will give you a maximum output of about 455 watts. That means that the wire gauge should be able to handle about 9 amps of current. Therefore, wire thickness will be 21 Ga. (.7mm)

I think that it will impossible to fit 72 coils with 250 turns each in the space that you have now.

Assuming that you manage to squeeze those coils in that space, there is bound to be some voltage cancellation because of the way the magnets are arranged. Therefore, much less than the maximum of 455 watts will be produced.

96 magnets of that size have the potential to produce 1,250 watts maximum. To get that much power, the coils would have to be placed between 2 magnets. That way the coil will be located where the maximum flux is.

With the way that the magnets are arranged in your drawing, less than 40% of the magnets flux will be reaching the coils.

Ed

[brandnewb]

I have some encouraging results.

I wound a single coil with 9 winds and spun the magnets at +- 60 rpm under it.

I am getting millivolt readings up to 134. These values serve as indicator only, not scientifically taken.

Now if the following math is applicable then I calculate I'll be needing to wind the following coils.

0.134 / 9 = 0.0148 volts per wind per coil.

target voltage 58.4 / 96 coils per phase (48 on each side per phase) = 0.608 volt per coil.

0.608 / 0.0148 = 41 winds per coil.

Now obviously I am not taking into account the law of diminishing returns as the winds get further and further from the magnets the more winds we wind. But with a target of 41 winds we can stay well within 4mm of the magnets. The area where my readings tell me that a repelling arrangement gives us more flux to play with than with a traditional arrangement.

Also not taking into account yet what will happen if the coils are swimming in 95% iron powder resin.

Does this make sense? If it does then we can return to the original question. What should I be aiming for 60 rpm or 120 rpm? I am leaning towards 60 because of the safety concerns.

[brandnewb]

I did not mean to say  swimming. baked in is a better term.
All the iron powder will also help with heat dissipation.

[MagnetJuice]

I'll check those calculations later. Something doesn't look right.

Iron in the stator? If you want to use this PMG for a VAWT, iron in the stator is not a good idea. A PMG for a VAWT should spin as freely as possible, that way it will start rotating with low winds. Iron introduces cogging (friction/resistance) and heat.

Also, you should think about this. Usually, the VAWT is designed first, dictated by the size that you can comfortably built and put in your property. Then you design a PMG to match that.

An H-Darrieus that can provide 1200 watts at 70 RPM at 9 m/s has a diameter of about 5 meters.

What is the biggest mill that you can put in your yard?

Ed

[brandnewb]

In the second link I posted in the original post in this thread I explain all about the turbine.
when using
https://www.omnicalculator.com/ecology/wind-turbine

I get an output power of 0.692KW at 9m/s
with a 2wx3h turbine

The average is less here though this is why I plan to have the turbine stack-able to create 2x6m

Local codes maxes the diam at 2m so I have to get power out height. it's too bad really this local code as torque comes more from diam than height.

I'll check those calculations later. Something doesn't look right.

Iron in the stator? If you want to use this PMG for a VAWT, iron in the stator is not a good idea. A PMG for a VAWT should spin as freely as possible, that way it will start rotating with low winds. Iron introduces cogging (friction/resistance) and heat.

Also, you should think about this. Usually, the VAWT is designed first, dictated by the size that you can comfortably built and put in your property. Then you design a PMG to match that.

An H-Darrieus that can provide 1200 watts at 70 RPM at 9 m/s has a diameter of about 5 meters.

What is the biggest mill that you can put in your yard?

Ed

[brandnewb]

regarding the cogging. I can also use the iron powder then to rotate along with the magnets on the outer parts of the PMG. A suggestion made by SparWeb if I remember correctly in another thread of mine.
I'll only use resin then for the coils to keep them from vibrating when under load.

[MagnetJuice]

I get an output power of 0.692KW at 9m/s
with a 2wx3h turbine

Yes, that sounds about right. But that power is generated at over 300 RPM, not at 70 RPM.

I recommend that you design the windmill first, then design the PMG to match it.

Ed

[brandnewb]

I am not sure what you mean. I get an output power of 0.692KW at 9m/s at 60 RPM.

I am probably am using the omnicalculator wrong but the values are correctly input.

Anyway what I really want to know is if 134 milli volt for a single 9 wind coil at 60 RPM is better than anyone expected or I should really go traditional to get even better results than that.

I get an output power of 0.692KW at 9m/s
with a 2wx3h turbine

Yes, that sounds about right. But that power is generated at over 300 RPM, not at 70 RPM.

I recommend that you design the windmill first, then design the PMG to match it.

Ed

[brandnewb]

Also do Magnetjuice and joestue seem to disagree on having a coil core out of iron powder?

The commercial PMG for a wind turbine that I bought did have an Iron Laminated core for the coils/stator as have most of the examples I have seen so far.

So if am a bit confused why it would not be good, or if having the outer powder disks rotate along with the magnets would be even better?

[MagnetJuice]

That online calculator doesn't allow for the user to choose RPM to be entered and considered in the calculations.

At the bottom, you can change the RPM but only so you can see the changes in TSR.

It seems that the online calculator is set to work with values close to .3 for solidity and somewhere between 2.8 and 3.8 for TSR.

You can lower the RPM of a machine by increasing the solidity. That lowers the TSR but it also lowers the Cp.

At 60 RPM the TSR for a 2 meter diameter machine is about .7, output power will be about zero at .7 TSR.

You can get about 3kW from a VAWT at 60 RPM, but the diameter will be about 7 meters. That will be a nice machine to be spinning on any backyard.

I think that joestue was talking about high speed machines. I agree with joestue that using a 30% iron PLA in the center of the coil is an idea worth experimenting with. The problem is finding a PLA that can withstand 180 C.

If the PMG is to be used on with a low RPM VAWT, in my opinion, no amount of ferromagnetic material should be used in the coils. Unless you have a lot of high winds.

Ed

[brandnewb]

I think I follow ever so slightly although I have not yet researched what all those abbreviations mean

But then there is the answer I was looking for.

I should be aiming for an rpm that matches the default rpm of that calculator yes?
Which is 137.5 for a VAWT and 180 for a HAWT if I want any realistic expectations based on that calculator yes?

If I did get that correctly then I think going for 137.5 rpm should be not a problem. I am only then a bit sad and upset that I have local codes (max diam) that are forcing me to sacrifice safety in order to be able to do any actual designing and power generation.

That online calculator doesn't allow for the user to choose RPM to be entered and considered in the calculations.

At the bottom, you can change the RPM but only so you can see the changes in TSR.

It seems that the online calculator is set to work with values close to .3 for solidity and somewhere between 2.8 and 3.8 for TSR.

You can lower the RPM of a machine by increasing the solidity. That lowers the TSR but it also lowers the Cp.

At 60 RPM the TSR for a 2 meter diameter machine is about .7, output power will be about zero at .7 TSR.

You can get about 3kW from a VAWT at 60 RPM, but the diameter will be about 7 meters. That will be a nice machine to be spinning on any backyard.

I think that joestue was talking about high speed machines. I agree with joestue that using a 30% iron PLA in the center of the coil is an idea worth experimenting with. The problem is finding a PLA that can withstand 180 C.

If the PMG is to be used on with a low RPM VAWT, in my opinion, no amount of ferromagnetic material should be used in the coils. Unless you have a lot of high winds.

Ed

[Adriaan Kragten]

This question has been asked already many times on this forum and has been answered by different people. Some months ago I have written an extended post on this forum called "Design of a PM-generator for a wind turbine" in which I have explained how the optimum number of turns per coil and the wire thickness have to be determined if the generator is coupled to a certain windmill rotor. But for optimal matching, you have to know the Cp-lambda curve of the windmill rotor and you have to be able to measure the mechanical and the electrical power of the generator for different loads. If you can't do that, it is try and error.

[MagnetJuice]

I made a mistake in my original calculations. I am sorry about that.

The number of turns per coil should be 32, not 250 as I stated previously.
Somehow I calculated for 3 coils per phase instead of 24 coils per phase.

My calculations were for only one stator with 72 coils on one side of the magnet rotor.

On the other hand, if you only get 134 millivolts with a coil of 9 turns, that tells me that the flux reaching the coils is slightly less than I calculated for. It also depends on how far the test coil was from the magnets.

As for the online calculator, I have no idea where you get the default RPM. As far as I know, there is no default RPM on that calculator. I don’t know where you got the 137.5 RPM.

They seem to have a default TSR value of 6, which is a bit low for a HAWT, but too high for an H-Darrieus VAWT.

Ed

[JW]

TSR is related to "tip speed rotation"

[brandnewb]

Yes thank you, that changes things considerably indeed.

And things would look even better if you would simulate using a traditional arrangement? As my data suggests that repelling would perform better than traditional if we can stay close to the magnets with the coils. And with that little winds that should be no problem.

Anyway I have decided I'll wait dabbling with FEMM until after I have an actual fully assembled generator and turbine prototype. So that I can be see if what I can get out of FEMM corresponds with what I am getting in practice.

So I will be making the prototype turbine now and will start with my current magnet disk and only straight symetrical airfoils.
As my charge controller keeps the circuit open until the cut-in voltage is reached I am hoping that there will be no issues with auto starting. Otherwise I think i'll try adding savonius drag type blades to the central column first to get it auto starting before I go the helix route. for auto start purposes. The reason for that is because when going helix I need 3 x 120 degrees as to make sure the helix blades are overlapping each other. And that sacrifices torque because then the straight blades can't extend all the way.

In terms of prototype scaling, would one think it's ok if I first build a 33% scaled version? that would mean 0.66m diam x 1m height and put it in a home made wind tunnel. Or simply on my lawn.
If that sounds like a good idea then shall I keep the blade chord length at 20cm or also scale that down to 33%

EDIT: Or the only real tests are done full scale?
I am going for 2x1 tests first. it will force me to think on the stackability of the turbine as well.

I made a mistake in my original calculations. I am sorry about that.

The number of turns per coil should be 32, not 250 as I stated previously.
Somehow I calculated for 3 coils per phase instead of 24 coils per phase.

My calculations were for only one stator with 72 coils on one side of the magnet rotor.

On the other hand, if you only get 134 millivolts with a coil of 9 turns, that tells me that the flux reaching the coils is slightly less than I calculated for. It also depends on how far the test coil was from the magnets.

As for the online calculator, I have no idea where you get the default RPM. As far as I know, there is no default RPM on that calculator. I don’t know where you got the 137.5 RPM.

They seem to have a default TSR value of 6, which is a bit low for a HAWT, but too high for an H-Darrieus VAWT.

Ed

[brandnewb]

TSR is related to "tip speed rotation"

hahha yes thank you. I realize now there weren't all that many different abbreviations as I had first thought I saw

[brandnewb]

As for the online calculator, I have no idea where you get the default RPM. As far as I know, there is no default RPM on that calculator. I don’t know where you got the 137.5 RPM.

They seem to have a default TSR value of 6, which is a bit low for a HAWT, but too high for an H-Darrieus VAWT.

Ed

Yes I also do not know. I checked again and now the default values are different than yesterday. Much higher.

Anyway I am going the trail and error route. So let's see where this leads us.

[brandnewb]

the more stable my tests get the better (or less bad) the results.

I am now at 177 millivolts at 60 rpm at a distance of 1.9mm from the magnets. My final goal is to have the stators and rotor touch at first and have friction remove some of the resin to end up with almost no distance between coils and magnets.

When applying the formula I used earlier I am now down to 31 winds per coil. I will see what happens when I spin at 120 rpm.

This is the test rig I will use for testing the generator iterations and the turbine iterations.