AXIAL FLUX 2

[xclr82xtc]

i think an illustration will help us all get on the same page...this is the design

im sorry you guys feel im not listening to you..i really am, i just dont get what your trying to say..

ok here is the specs for this

coil length 2 inches

turns 80

1x1 inch coils

4 layers of turns

each magnet is rated at .6403 tesla..i can give you the link if you want

the center large magnet is diametrically alligned so that it will stabalize the horizontal axis of the mill...not nessacary for the PMA but i thought i would show you anyways.

come on guys...i just went and spent 60 dollars on physics books so i can get this...i really want to understand.

NOTE..THE ACTUAL MODEL WILL HAVE 21 COILS...7 Per PHASE...but i feel this representation is the simplest way to get the idea across.  i know that i dont have a number divisible by 3 right now (for 3 phase)..but you get the idea i hope

[xclr82xtc]

bigger

[xclr82xtc]

ok last one...the poles were wrong..didnt realize. i reversed them in this drawing

http://www.fieldlines.com/images/scimages/15777/levmag.jpg

[wooferhound]

is it possible to get a Disc magnet that is polarized to the center like that ?

[Hilltopgrange]

Unless I completely misunderstand your drawing this is a radial flux you are suggesting and not an axial flux as stated! you would be better to double the diameter and put all your mags to the outside with a single stator, with an iron backing ring for the flux return path.  Still wont work as a vawt though! There was a contest on this site a while back to produce a working vawt that actually produced some power, as far as I can remember nobody claimed the prize. That should tell you something in itself.

hilltopgrange

[xclr82xtc]

yes it is..they make them.

[xclr82xtc]

from my calculations...

34 inch diamter centerline of the coil...it would be .05615 seconds for 1 turn. the large radius of the machine will create enough speed, where as with a wind turbine you have huge blade but the PMA is only a fraction of the circumference of the blades..even though the VAWT turns relativley slowly compared to a HAWT, it produces more torque so you can use larger magnets and not have to worry about cogging.

[xclr82xtc]

guess i should build this and try to make some power huh? how much was the prize lol?

[frepdx]

Hi xclr82xtc,

I didn't read through the other thread because there were some unpleasant posts, so forgive me if this has already been covered, but just about everything you need for a PMA design is here; http://www.fieldlines.com/story/2004/10/15/22250/879

read down to the posts from Hugh Piggott;

"Turns per phase = 17000 x (System voltage V +1.4) / ( Desired cut in speed RPM) / (Strength of flux in Tesla) / (total area of magnet face in square inches)"

Some rules of thumb - if you have 1/2 inch thick magnets, try to keep the gap between them (where the coil is) not much more than 1/2 inch. Design a completed magnetic circuit; for example, if you have a ring of magnets N,S,N,S try to mount them on a steel ring so their magnetic flux is carried to each other. KISS (you should be familiar with that on

[xclr82xtc]

thank you.

i need to forumlas for a design i already have...i know how many turns, and coils. im trying to figure out the output. it looks to me that formula tells you how many turns you need. for a ceratin voltage.is that right?

[frepdx]

Yes, that's right. And you also need to calculate the coil resistance and make an educated guess about open circuit voltage and voltage under load.

[xclr82xtc]

THATS WHAT I WAS DOING IN THE OTHER THREAD AND EVERY ONE SAID I WAS WRONG!!!

WTF?

sorry not directed at you frep...

i did the calculations using the faraday forumla. and everyone said they were wrong..can you help?

[xclr82xtc]

coil info

1x1x2

80 turns total, 4 rows of twenty.

center line of coil is 34 inches.

the magnets making 1 revolution per minute would pass the coils at X coils per second...lets figure out X

circumference of a 34 inch circle = 106.8 inches

divide that by 60 seconds...106.8/60= 2.68

so EVERY SECOND...the magnets travel 2.68 inches.  in turn, the magnetic fields pass through the coils at the rate of 2.68inches per second?

am i right to this point?

[xclr82xtc]

there is 80 coils in 2 inches, so in 2.68 we would have 107turns

the tesla of the magnets is .6403

the area is .75inchx2inch then change to meters squared = 0.003610meters squared

faradays law

N = -1 * (-V/ (( tesla * area meters squared)/ seconds))

107=-1*(-V/((.6403*.003610)2.68)

107=-1*(-V.002311)

-107=(-V/.002311)

.2472=V

so right now we have 107 turns = .2472 v at at rate of 1 rpm....

now there is 7 coils of 80 so thats 560 total turns

divide by 107 =5.233

so times 5.233  time the Voltage .2472

=1.29 volts per phase...

my math looks more right this time lol but is it?

[xclr82xtc]

now how many feet of wire in a coil (we need to find the resistance)

wire info

Nominal Diameter: 0.10 inch

Nominal OHMS/LB. -20 Celsius: 0.0314

Feet/LB: 31.48

outersize of coil is 1 inch. 4 layers and the core is .6inches.(yes this does take into account wire size) .so the  perimter for one layer of turns  is (.7*4)+(.8*4)+(.9*4)+.(1*4)2.8+3.2+3.6+4=13.6inches


now the coil is 2 inches long so we need have 20 layers...20*13.6=272inches...divde by 12=22.6 feet or wire per coil...nice. lol


ok so now we know we have 7 coils so 22.6*7 = 158.2 feet.


its 31.48 feet per lb  so 158.2/31.48=5.025lbs


5.025 * .0314.1578

so the TOTAL resistance of all seven coils is .1578ohms

[Perry1]

What does 106.8/60 equal again?

[xclr82xtc]

1.78..

did i mess up? lol

let me check it again...i got more path i need to post

[xclr82xtc]

damn i fat finger it...lol..that messed it all up

ill go back and redo it. just dont post anything, ill post all the math when im done.

[xclr82xtc]

ok here is all the math...i think its right..feel free to check it, i am at the point where i need to  convert my single phase numbers to 3 phase, and acccount for having 15 pairs of magnets...not just 1, which is what these calcs show.

i think with three phase you leave the voltage the same and times the amperage by 70.7% is that right? i saw that somewhere last night, but i dont know if its right.[b]

coil info

1x1x2

80 turns total, 4 rows of twenty.

center line of coil is 34 inches.

the magnets making 1 revolution per minute would pass the coils at X coils per second...lets figure out X

circumference of a 34 inch circle = 106.8 inches

divide that by 60 seconds...106.8/60= 1.78

so EVERY SECOND...the magnets travel 1.78 inches.  in turn, the magnetic fields pass through the coils at the rate of 1.78 inches per second?

am i right to this point?

____________

there is 80 coils in 2 inches, so in 1.78seconds we would have  

the tesla of the magnets is .6403

the area is .75inchx2inch then change to meters squared = 0.003610meters squared

faradays law

N = -1 * (-V (( tesla * area meters squared)/ seconds))

71.2=-1*(-V/((.6403*.003610)1.78)

71.2=-1*(-V)

-71.2=(-V/.00129)

.0918=V

so right now we have 71.2 turns = ..0918 v at at rate of 1 rpm....

now there is 7 coils of 80 so thats 560 total turns

divide by 71.2 7.87

so times 7.87  time the Voltage .0918


=.722 volts per phase...


my math looks more right this time lol

--------------------------------------------------------------


now how many feet of wire in a coil (we need to find the resistance)


wire info


Nominal Diameter: 0.10 inch


Nominal OHMS/LB. -20 Celsius: 0.0314


Feet/LB: 31.48


outersize of coil is 1 inch. 4 layers and the core is .6inches.(yes this does take into account wire size) .so the  perimter for one layer of turns  is (.7*4)+(.8*4)+(.9*4)+.(1*4)=2.8+3.2+3.6+4=13.6inches


now the coil is 2 inches long so we need have 20 layers...20*13.6=272inches...divde by 12=22.6 feet or wire per coil...nice. lol


ok so now we know we have 7 coils so 22.6*7 = 158.2 feet.


its 31.48 feet per lb  so 158.2/31.48=5.025lbs


5.025 * .0314.1578

so the TOTAL resistance of all seven coils is .1578ohms

__________________

so now we know the total voltage of one phase at 1 rpm of the mill, and we know the constant resistance of the coil. (i also have the inductance of each coil but i dont think we need it.)

so with this info, we SHOULD be able to figure out the amps using ohms law. i=v/r

.722/.1578=4.578AMPS

SOOOO......per phase, at 1 rpm we have 1.29 volts and 4.578 amps. kinda crappy, but hey its only turning at 1 rpm, and thats only using 1 pair of magnets, there is 15 total pairs.

[xclr82xtc]

there is 80 coils in 2 inches, so in 1.78seconds we would have  

this line should read

there is 80 coils in 2 inches, so in 1.78inchs we would have  71.2coils

[xclr82xtc]

using this caLC

http://www.csgnetwork.com/ohmslaw2.html

I  get 3.303 watts per phase.times that by 3 phases and i get 9.909watts

enter the volts and watts into the 3 phase calc and i get roughly 8 amps...

However i have no clue if this is right.

times that by 15 magnet pairs and wouldnt you get 40 amps at the speed of 1rpm?

it might stall but is my math at least right?

[Perry1]

Are your magnets or rectangular? Can you be more specific on their size and shape?

What do your coils look like? Short fat cylinders or long thin. Can you eloborate on their size and exact shape?

Perry

[xclr82xtc]

the picture at the top of this page show the setup..so you can get what im talking about..

the magnets are cylinder .75inch diameter 2 inches long....here ...its these big boys

http://www.kjmagnetics.com/proddetail.asp?prod=DCY0

i used the full gauss amount for the calculations since im actually using two magnets. realistically, the number should probably be closer to .8 tesla

the Coils are basically 2inch cubes stretched out to a 2 inch rectangle.

i was right on at 80 turns using a .6 inch air core, it just happed to work out to be 2 inches.. that was just a good guess though lol.

the coils are wrapped in 20 turn layers, 4 layers for a total of 80 turns

[xclr82xtc]

sorry i describe the coils wrong...they are basicall TWO 1 inchcubes sat side by said to form a 1x1x2 rectangle. sorry my other description was misleading

[Hilltopgrange]

1. Axial flux alts as built here do NOT cog, they cant cog if they don't have iron in the stator
   
   
2. 15 magnet poles wont work as the poles must alternate ie NSNSNS YOU Must use an even number of poles otherwise you end up with two like poles beside each other ie NSNSN
   
   
3. The ratio of coils to mags should be 4 poles to 3 coils
   
   
4. when you know the output voltage of a single coil you multiply that figure by the number of coils per phase and then you multiply that by 1.73 for 3 phase.

   
   
   

You should spend some time reading past stories on this site as it has all been covered many times and will save you a lot of time making very basic mistakes.

A more simple approach is to build your magnet rotors and then insert a test coil same as I did on my last build here

Part one http://www.fieldlines.com/story/2009/5/23/135436/243

Part two http://www.fieldlines.com/story/2009/6/21/22551/6284

Part three http://www.fieldlines.com/story/2009/7/20/11923/2353

going to bed its 3.30 am here

hilltopgrange

[xclr82xtc]

thanks alot hilltop, those threads looklike they will help alot.

when you have some time, i hope maybe you can read more into my calcualtions and help my with the 3 phase conversions.

[Perry1]

A couple places where you are veering off course.

1.) Your magnets are rods, why are you calculating their area like they are a rectangle? Us pi*r^2 for the area.

2.) for your time in the denom (seconds) you use 1.72 inches per second. Is that a unit of time? 'speed' in the formula refers to how many seconds it takes to perform a revolution. you should be using 60 sec here. Always put units in your calc's, if they don't cancel out in the end you have a wrong answer. It's usually a bad sign when your equation has both english and metric units in it.

3.) The Gauss you are quoting for this magnet is surface Gauss. Near the center of your air gap it would be approx 50%

Perry

[xclr82xtc]

1 i used a cylinder surface area calulator. i will do it by hand

2 i was told that the N should equal the seconds when you didnt know the voltage. likewise if you knew the voltage and didnt know the N then you would of course use 60 seconds. but i will redo it, and see what happens. i still have not competly grasped faradays law but i will try

3 agree that it will drop by half. but using 2 magents would bring it back up wouldnt it?

[xclr82xtc]

Surface Area = 2(pi r 2) + (2 pi r)* h

that answer was correct for surface area

N = -1 * (-V/ (( tesla * area meters squared)/ seconds))

71.2=-1*(-v/((6304*0.003610)60)

71.2 -1*(-v((.002276/60)


71.2 =-1*(-v/((3.75*10^-5)


-72.1(v/.0000375)

.00541=v

look right?

[Perry1]

Area is = pi * r^2

You only want the surface area that the flux passes through perpendicularly.

[xclr82xtc]

ah...ok, that website told me entire area, if thats the case, i need to get magnets with a larger radius