Author Topic: Axial Flux Designer  (Read 23025 times)

0 Members and 1 Guest are viewing this topic.

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Axial Flux Designer
« on: September 21, 2015, 04:33:19 PM »
Hello,

This post will be used to discuss the Axial Flux designer.

This version (see attachement) is mainly to get a first impression of what the program looks like. I have made a lot of unit tests. However, there is still some untested code!

Keep in mind that there are still many improvements to be made (iterator is not implemented)! Consider it a preview version :)

I am mainly in the process of doing proper documentation of the code and writing a small manual about the used formulas and how to use the software.

Hopefully I can put the link of the github repositry and documentation online.



DamonHD: .exe removed at request
« Last Edit: September 28, 2015, 05:02:27 AM by DamonHD »

DamonHD

  • Administrator
  • Super Hero Member Plus
  • *****
  • Posts: 4125
  • Country: gb
    • Earth Notes
Re: Axial Flux Designer
« Reply #1 on: September 21, 2015, 04:52:11 PM »
Everyone will be sure to let their favourite virus checker look at the .exe before running it of course!

May be good to reiterate system requirements.

Rgds

Damon
Podcast: https://www.earth.org.uk/SECTION_podcast.html

@DamonHD@mastodon.social

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #2 on: September 21, 2015, 05:00:17 PM »
This is the link to the github repositry:

https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/tree/master/AxialFluxGeneratorDesigner

This will be the place were all source code will be posted.


kc7noa

  • Newbie
  • *
  • Posts: 39
  • Country: us
Re: Axial Flux Designer
« Reply #3 on: September 21, 2015, 05:58:54 PM »
What program runs a ".cs" file extension ?

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #4 on: September 21, 2015, 06:14:10 PM »

boB

  • Sr. Member
  • ****
  • Posts: 389
  • Country: us
    • boB
Re: Axial Flux Designer
« Reply #5 on: September 21, 2015, 10:40:15 PM »

VS 2013 does not seem to like the "="  initializer

eg.
  public double TurbineSpeedTipRatioMax { get; set; } = 7;

boB

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #6 on: September 22, 2015, 12:42:53 AM »
Setting a default value of a property is a new feature of the .net framework that is only supported in VS 2015.

The file that is in the repository is only the  generator class! The gui files are not available yet!

boB

  • Sr. Member
  • ****
  • Posts: 389
  • Country: us
    • boB
Re: Axial Flux Designer
« Reply #7 on: September 22, 2015, 01:34:46 AM »
Setting a default value of a property is a new feature of the .net framework that is only supported in VS 2015.

The file that is in the repository is only the  generator class! The gui files are not available yet!

Thanks.  I suppose the variables could be initialized separately then.

Or, how about compiling in a legacy non- dot-net mode ?   I really don't care for dot-net when normal C or C++ could work fine.  Makes the programs a bit more portable... maybe.  Or maybe not.

Will wait for Oooie GUI stuff.

boB

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #8 on: September 22, 2015, 03:09:13 AM »
Well, I use C# as a programming language :) Compiling to C++ is possivble through code conversion but why?? C# is the successor of c and c++.

C# code can be compiled into an executable so portability is not an issue.


rsskarthikeyan

  • Newbie
  • *
  • Posts: 15
  • Country: in
Re: Axial Flux Designer
« Reply #9 on: September 24, 2015, 03:28:18 PM »
anyone have the solution for this article. is this article worth to follow ?
Axial Flux Permanent Magnet Generator Design for Low Cost Manufacturing of Small Wind Turbines by
K.C. Latoufis, G.M. Messinis, P.C. Kotsampopoulos and N.D. Hatziargyriou.



StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #10 on: September 24, 2015, 03:46:49 PM »
The article is absolutely worth following. Your Excel sheet looks ok. However, some data is not set correct. I advise you to do the calculations one step at a time! If you have doubts you can ask them here. But I am not going to correct your Excel sheet (I just don't have the time)...

In the first pages of the article a list of symbols is available (including correct units). I've noticed that you use a lot of mm instead of meters. Also the case study (chapter 8 ) can be used to check the calculations.

If you have some basic programming experience (C# in special) you can check the generator class I made based on this article.

https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/blob/master/AxialFluxGeneratorDesigner/AfpmCalculations.cs

It should give you some more insight on he calculations that are done.

Regards.
« Last Edit: September 24, 2015, 03:59:00 PM by StirlingEngine »

boB

  • Sr. Member
  • ****
  • Posts: 389
  • Country: us
    • boB
Re: Axial Flux Designer
« Reply #11 on: September 25, 2015, 02:18:33 PM »
Well, I use C# as a programming language :) Compiling to C++ is possivble through code conversion but why?? C# is the successor of c and c++.

C# code can be compiled into an executable so portability is not an issue.


Does OS X (unix) or Linux have a C# compiler ?

boB

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #12 on: September 25, 2015, 02:56:01 PM »
OSX and Linux support C# using mono.

http://www.mono-project.com/
http://www.mono-project.com/docs/about-mono/supported-platforms/

I also using on my Linux machine and works great :)

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #13 on: September 25, 2015, 07:12:53 PM »
Tomorrow I will put a beta version of the executable on github. This version contains the designer and the iterator.

https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/tree/master/AxialFluxGeneratorDesigner/bin/Debug

From then I will start working on:
1. The possibility to chart iterated data
2. Tutorial.

If people are intrested I will also upload the rest of the project (GUI files) so people can build/modify the software on a local pc.

Regards :)

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #14 on: September 27, 2015, 02:11:27 PM »
I've put an Alpha version online.

https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/tree/master/AxialFluxGeneratorDesigner/bin/Debug

This version contains:
1. The designer
2. The iterator
3. Data chart

From now I will mainly start working on the documentation and a tutorial. Hopefully from there we can discuss the calculations and make improvements.

Regards :)

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #15 on: September 28, 2015, 04:52:48 PM »

rsskarthikeyan

  • Newbie
  • *
  • Posts: 15
  • Country: in
Re: Axial Flux Designer
« Reply #16 on: September 29, 2015, 06:33:19 AM »
Hello,

Today I've started documenting the project :) (well first I have to learn LaTex https://en.wikipedia.org/wiki/LaTeX)

Anyway, here is the link where the output will be!

Folder: https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/blob/master/Documentation
File: https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/blob/master/Documentation/AfpmgDesignerDocumentation.pdf

Regards.
[                     ]

I could not run your software. kindly upload correct version to use others. thank you

rsskarthikeyan

  • Newbie
  • *
  • Posts: 15
  • Country: in
Re: Axial Flux Designer
« Reply #17 on: September 29, 2015, 08:31:26 AM »
The article is absolutely worth following. Your Excel sheet looks ok. However, some data is not set correct. I advise you to do the calculations one step at a time! If you have doubts you can ask them here. But I am not going to correct your Excel sheet (I just don't have the time)...

In the first pages of the article a list of symbols is available (including correct units). I've noticed that you use a lot of mm instead of meters. Also the case study (chapter 8 ) can be used to check the calculations.

If you have some basic programming experience (C# in special) you can check the generator class I made based on this article.

https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/blob/master/AxialFluxGeneratorDesigner/AfpmCalculations.cs

It should give you some more insight on he calculations that are done.

Regards.

I have one doubts, i incorporated your suggestion and finally i get Wc value 27mm actual value on this paper is
Wc (0.313786028)? kindly clarify my doubt.

Thank you

rsskarthikeyan

  • Newbie
  • *
  • Posts: 15
  • Country: in
Re: Axial Flux Designer
« Reply #18 on: September 29, 2015, 12:16:08 PM »
The article is absolutely worth following. Your Excel sheet looks ok. However, some data is not set correct. I advise you to do the calculations one step at a time! If you have doubts you can ask them here. But I am not going to correct your Excel sheet (I just don't have the time)...

In the first pages of the article a list of symbols is available (including correct units). I've noticed that you use a lot of mm instead of meters. Also the case study (chapter 8 ) can be used to check the calculations.

If you have some basic programming experience (C# in special) you can check the generator class I made based on this article.

https://github.com/FinalFrontierPrototyping/AxialFluxGeneratorDesigner/blob/master/AxialFluxGeneratorDesigner/AfpmCalculations.cs

It should give you some more insight on he calculations that are done.

Regards.

i have incorporated your suggestion in my calculation. but i could not get the values of  width of coil Wc 31.54 and getting only 27mm.....have you got the all values as it is in that article.

Mary B

  • Administrator
  • SuperHero Member
  • *****
  • Posts: 3169
Re: Axial Flux Designer
« Reply #19 on: September 29, 2015, 06:50:23 PM »
I just ran it and your rotor diameters are not real world. with 1x2x.5 inch thick(yes I converted to mm) magnets it was telling me a 100mm diameter rotor. Not going to happen! 10 inch rotor is about typical for that!

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #20 on: September 30, 2015, 03:38:56 AM »
Hello Mary B,

I do all the calculations based on an article http://www.academia.edu/7610343/Axial_Flux_Permanent_Magnet_Generator_Design_for_Low_Cost_Manufacturing_of_Small_Wind. The calculation for the rotor radius is as follows:



rho = 1.2
Rturbine is the turbine radius (m)
Vaverage is the average/optimal wind speed (m/s)
Cp = power coefficient = 0.35
eta = generator efficiency = 0.9

The rotor radius calculated in this article is not related to the magnet size. What calculations do you use?

I am working on the documentation of the sowfware. This document will describe all the used formula's :)

Regards.
 
« Last Edit: September 30, 2015, 03:45:31 AM by StirlingEngine »

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #21 on: September 30, 2015, 06:14:16 AM »


i have incorporated your suggestion in my calculation. but i could not get the values of  width of coil Wc 31.54 and getting only 27mm.....have you got the all values as it is in that article.
[/quote]

Hello maybe the outcome of the formula is cm instead of m? I have emailed the author of the article about this.

The difference between 27 and 31 mm is probably due to a different  Coil fill factor (Kf). Page 419 describes that the Kf factor can vary due to the accuracy of the coil winding process. For example if you use a Kf of 0.42 instead of 0.55 then you get 31.67 mm.

http://www.coilgun.eclipse.co.uk/coil_resistance_formula.html

Regards.

rsskarthikeyan

  • Newbie
  • *
  • Posts: 15
  • Country: in
Re: Axial Flux Designer
« Reply #22 on: September 30, 2015, 12:16:13 PM »


i have incorporated your suggestion in my calculation. but i could not get the values of  width of coil Wc 31.54 and getting only 27mm.....have you got the all values as it is in that article.

Hello maybe the outcome of the formula is cm instead of m? I have emailed the author of the article about this.

The difference between 27 and 31 mm is probably due to a different  Coil fill factor (Kf). Page 419 describes that the Kf factor can vary due to the accuracy of the coil winding process. For example if you use a Kf of 0.42 instead of 0.55 then you get 31.67 mm.

http://www.coilgun.eclipse.co.uk/coil_resistance_formula.html

Regards.
[/quote]

i already mail them one month before. Not yet get reply from them... I'm going to fabricate the generator as per this paper

Mary B

  • Administrator
  • SuperHero Member
  • *****
  • Posts: 3169
Re: Axial Flux Designer
« Reply #23 on: September 30, 2015, 07:57:34 PM »
Just real world observations that you will not get that many magnets into that tiny space! Same for coils! From what I have read the coil center hoe should be the same size as the magnet.

Hello Mary B,

I do all the calculations based on an article http://www.academia.edu/7610343/Axial_Flux_Permanent_Magnet_Generator_Design_for_Low_Cost_Manufacturing_of_Small_Wind. The calculation for the rotor radius is as follows:

(Attachment Link)

rho = 1.2
Rturbine is the turbine radius (m)
Vaverage is the average/optimal wind speed (m/s)
Cp = power coefficient = 0.35
eta = generator efficiency = 0.9

The rotor radius calculated in this article is not related to the magnet size. What calculations do you use?

I am working on the documentation of the sowfware. This document will describe all the used formula's :)

Regards.

joestue

  • Hero Member
  • *****
  • Posts: 1759
  • Country: 00
Re: Axial Flux Designer
« Reply #24 on: September 30, 2015, 08:42:33 PM »
Just real world observations that you will not get that many magnets into that tiny space! Same for coils! From what I have read the coil center hoe should be the same size as the magnet.

no, ideally the magnets are pie shaped, like a trapezoid but rounded on the ends and middle.

say you have 12 magnets, each magnet could theoretically take up 30 degrees. 90 perent of this is 27 degrees and that would be about ideal.
My wife says I'm not just a different colored rubik's cube, i am a rubik's knot in a cage.

Mary B

  • Administrator
  • SuperHero Member
  • *****
  • Posts: 3169
Re: Axial Flux Designer
« Reply #25 on: October 01, 2015, 07:22:14 PM »
Most commonly available magnets are rectangular though...

joestue

  • Hero Member
  • *****
  • Posts: 1759
  • Country: 00
Re: Axial Flux Designer
« Reply #26 on: October 02, 2015, 01:12:11 AM »
Most commonly available magnets are rectangular though...

Far as I can tell given the rectangular 2:1 ratio limitation, the inside corners should touch each other, or perhaps even shave the corners off to push them closer together, depending on the pole count. if the pole count is nearly infinite, they should be spaced about 1/8th of their width, apart. this fits the standard 150/180 degree arc that nearly every electrical machine uses. while the exact arc is selected to minimize cogging for iron cores, anything significantly less reduces the voltage.

However, given copper is one third the price per cubic inch of neodymium magnet.. there may be some marginal gain to be made by increasing the diameter. but the only reason i can come up with is because it increases the surface area available for cooling the coil.*

I have said this years ago, but the coils should be overlapping, trapezoidal and pressed. thanks to three-d printers you could print coil forms to wind the coils on, but you can make them from wood too. when you have an overlapping coil set, you get nearly double the copper density compared to discrete single coils spaced out like donuts on a tray.

with the large inside to outside diameters made by using 2 or 3 inch long magnets and >12 inch diameter disks, you have no shortage of space on the ID for the end turns of the copper coils, unlike the iron core examples where the  optimal figures are more like 12 inch diameter disk, and 4 inch long pie shaped magnets.


*and i'm not sure this is helpful either because reduced emf and higher harmonic content only makes matters worse.
on the plus side however, more copper equals more thermal mass. so the greater thermal mass may offset reduced efficiency.
My wife says I'm not just a different colored rubik's cube, i am a rubik's knot in a cage.

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #27 on: October 05, 2015, 03:32:18 AM »

i have incorporated your suggestion in my calculation. but i could not get the values of  width of coil Wc 31.54 and getting only 27mm.....have you got the all values as it is in that article.

Hello maybe the outcome of the formula is cm instead of m? I have emailed the author of the article about this.

The difference between 27 and 31 mm is probably due to a different  Coil fill factor (Kf). Page 419 describes that the Kf factor can vary due to the accuracy of the coil winding process. For example if you use a Kf of 0.42 instead of 0.55 then you get 31.67 mm.

http://www.coilgun.eclipse.co.uk/coil_resistance_formula.html

Regards.

Please note that the Heat dissipation value of 0.3 W/cm2 should be converted to its correct unit W/m2! so 0.3 W/cm2 = 3000 W/m2. If this value is used the correct Wc value is obtained.  I think that the author of the article started with a initial value of 0.55 for the Kf value. However, adjusted this value after winding the coils.

Regards.

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #28 on: October 05, 2015, 03:45:02 AM »
I just ran it and your rotor diameters are not real world. with 1x2x.5 inch thick(yes I converted to mm) magnets it was telling me a 100mm diameter rotor. Not going to happen! 10 inch rotor is about typical for that!

Hello Mary B,

I misunderstood the meaning of the rotor radius. I thought you mentionded the wind turbine rotor......   :-[

The article uses another formula (see chapter 6 of the article). I used a modified version, I will change it back so you can compare it.

Regards.

rsskarthikeyan

  • Newbie
  • *
  • Posts: 15
  • Country: in
Re: Axial Flux Designer
« Reply #29 on: October 05, 2015, 04:01:51 AM »

i have incorporated your suggestion in my calculation. but i could not get the values of  width of coil Wc 31.54 and getting only 27mm.....have you got the all values as it is in that article.

Hello maybe the outcome of the formula is cm instead of m? I have emailed the author of the article about this.

The difference between 27 and 31 mm is probably due to a different  Coil fill factor (Kf). Page 419 describes that the Kf factor can vary due to the accuracy of the coil winding process. For example if you use a Kf of 0.42 instead of 0.55 then you get 31.67 mm.

http://www.coilgun.eclipse.co.uk/coil_resistance_formula.html

Regards.

Please note that the Heat dissipation value of 0.3 W/cm2 should be converted to its correct unit W/m2! so 0.3 W/cm2 = 3000 W/m2. If this value is used the correct Wc value is obtained.  I think that the author of the article started with a initial value of 0.55 for the Kf value. However, adjusted this value after winding the coils.

Regards.

Yes i have used 3000W/m2 in my calculation..... i think changes made in winding coil value

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #30 on: October 09, 2015, 06:33:43 PM »
Hello,

When I read the article I miss some requirements regarding the coil dimensions vs. magnet dimensions.

How do you usually dimension your coil vs. magnet? I have seen magnets positioned it the length and magnet being positioned in the width?

Also should a coil have  pointed or stump end? I know a pointed coil has a larger copper surface. However, what are the advantages/disadvantages?

I would like to use a "rule of thumb" regardeing the coil vs. magnet size.

What are your experiences?

Regards,

JR.

StirlingEngine

  • Newbie
  • *
  • Posts: 32
  • Country: nl
Re: Axial Flux Designer
« Reply #31 on: October 15, 2015, 04:50:19 PM »
anyone with some advice?

Well, I continued :) I have been working on a concept that dynamicially calculates and draws!! the coils (complete rotor actually). This is done using FreeCAD and Python.

I am now in the process making the designer call the python script so the generated images are displayed.

I think that eventually a full 3D model of the designed AFPMG will be displayed :)

I will now continue with the implementation and the documentation (had a writers block ;) )

@MaryB can you provide me with some practical data? In this way I can test the code.
« Last Edit: October 15, 2015, 05:00:17 PM by StirlingEngine »

Mary B

  • Administrator
  • SuperHero Member
  • *****
  • Posts: 3169
Re: Axial Flux Designer
« Reply #32 on: October 16, 2015, 05:55:13 PM »
Look at the common designs here, and on otherpower. http://www.otherpower.com/turbineplans.html