Author Topic: An Alternate PM Alternator Design  (Read 17323 times)

0 Members and 1 Guest are viewing this topic.

jimovonz

  • Sr. Member
  • ****
  • Posts: 339
Re: An Alternate PM Alternator Design
« Reply #33 on: May 21, 2007, 01:40:27 PM »
GeoM, you have mentioned a few times recently about RMS voltage measurements and how they relate only to sine waves and not to our typical alternators. This is not my understanding. I believe RMS measurements can apply (and have meaning) to any wave form. Cheaper multimeters expect to see a sine wave and approximate an RMS measurement based on this assumption (0.707 x Vpeak for instance). Meters that are 'True RMS' calculate the RMS value based on the actual waveform measured. I believe that the RMS value gives the 'equivalent DC heating power' i.e. a 10Vrms waveform will dissipate the same heat across a given resistor as 10Vdc. I often see large variations in reading between my cheap meters and those that are 'True RMS' I only ever give any credit to the 'True RMS' meters when such discrepancies arise.
« Last Edit: May 21, 2007, 01:40:27 PM by jimovonz »

Lumberjack

  • Jr. Member
  • **
  • Posts: 88
Re: An Alternate PM Alternator Design
« Reply #34 on: May 21, 2007, 10:06:14 PM »
Didnt mean to shout....


I will go through your design again but I dont see anything new in it.Looks fairly close to the normal setup used around here. ( or was that the point?)


On RMS:

As you said RMS really only occurs in true sine waves but since most meters are built to read RMS and most here only have this type of meter it becomes an issue for most of us. This is a differance between theory and the real world. Yes the reading we get is incorrect but it is still refered to as RMS and not peak.


"Coils never reach "saturation".  They can't.  They do not contain any iron."


Here is a point where you are dead wrong. Any object that is passed through a magnetic field builds up flux to a maximum amount. This is a saturation point. Iron has nothing to do with it. Even a wood block undergoes this process and becomes saturated. All it takes is time spent inside a magnetic field.


"Nor are the speeds reached by these alternators high enough for hysteresis or inductive effects to become important."


These effect are governed by the "rate of change". Two magnets of opposite poles butted against each other create a very steep rate of change. You should also realise that a 60 hertz square wave or sawtooth wave can create harmonics well into the 100 khz range.


The semantics you dont care about killed a co-worker two years ago....All it takes is .1 amp at 5 volts to stop a human heart. A lot of the generators here are 3 phase and the neutrals are not always grounded.


Datum node ... nope never heard of it outside of programming.

« Last Edit: May 21, 2007, 10:06:14 PM by Lumberjack »

finnsawyer

  • Hero Member
  • *****
  • Posts: 1565
Re: An Alternate PM Alternator Design
« Reply #35 on: May 23, 2007, 09:26:16 AM »
"Here is a point where you are dead wrong. Any object that is passed through a magnetic field builds up flux to a maximum amount."


You're showing your lack of understanding of the physics or more precisely Faraday's Law, which states that the voltage induced in a loop of wire is equal to the time rate of change of the magnetic flux.  While the flux may reach the same maximum value, which does not constitute saturation, as increasing the magnet's strength will result in more flux, the case where the coil is moving faster will induce a larger voltage or a voltage pulse having a greater peak voltage.  If you consider the standard 12 magnet 9 coil three phase design, you find that the a coil will move 30 degrees in creating a single pulse (from magnet center to following opposite polarity magnet center).  With my proposed design Two groups of coils act together.  When the group one coils are centered over a magnet at zero degrees, the centers of the next group of coils are advanced 20 degrees, since there are eighteen coils. The coils of group two are just touching the next opposite pole group of magnets, the centers of which are advanced 30 degrees from zero.  As the group one coils move off of their magnets the second group coils move onto magnets having an opposite pole.  Their voltages add.  They must move only 10 degrees to produce the pulse.  If they move the ten degrees in the same time as the three phase coils move thirty degrees, the combined voltage of a coil from each of the two groups will create an identical pulse of voltage as a single coil of the three phase.  If the alternators have the same RPM, then each coil pair of my design will have a peak voltage three time that of the single coil of the three phase as they are moving three times as fast.  There will be six such pairs acting at any time, so we could write the peak voltage as 18V.  In the case of the three phase alternator there are three coils per phase.  So each phase will put out a peak voltage of 3V.  In the star configuration two phases are active, but are 120 degrees out of phase, taking into account the resulting 1.73 factor we find the peak voltage will be 5.19V.  So, we find that my design will have a peak voltage 3.47 times as great as any lead pair of the star connected three phase.  It also will have a resistance tree times as great for identical coils (18 coils in series versus six).  When connecting the alternators to matched resistances, you find that my design can output 4 times the power for the same number and size of magnets.  That said, this is a theoretical result. Other factors, such as rectification and application will affect the result.


A couple of points about harmonics:  Even for a square wave the energy in the harmonics dies out fairly rapidly.  The coils do have some inductance, so the harmonics are suppressed fairly quickly with frequency.  That is, the waveform will not deviate too much from a sine wave.


Saturation implies a maximum physical limit beyond which one can not go.  Iron has a magnetic saturation point.  Air, copper, or free space does not.  If you can create it, you can ram as much magnetic flux through a coil as you please.  There is no saturation point.  You are misusing the term to refer to the maximum value of flux that passes through the coil for a given construction, which depends on a number of factors that can be changed.

« Last Edit: May 23, 2007, 09:26:16 AM by finnsawyer »

Lumberjack

  • Jr. Member
  • **
  • Posts: 88
Re: An Alternate PM Alternator Design
« Reply #36 on: May 23, 2007, 07:01:10 PM »
You are right in that I was mis-using the term saturation however, given that the rotor / stator combination in use cannot be changed while in use my point is still valid. There is a maximum amount of flux that can be created at a given speed and once that it is reached the coil will stop producing. I can not find any term for that point and have always heard the term saturation used. Perhaps you might know a correct term for it?
« Last Edit: May 23, 2007, 07:01:10 PM by Lumberjack »

finnsawyer

  • Hero Member
  • *****
  • Posts: 1565
Re: An Alternate PM Alternator Design
« Reply #37 on: May 24, 2007, 08:47:00 AM »
Simply refer to it as the value of magnetic flux for that alternator design or produced by the given magnets.  Or use the time rate of change of the flux through the coils, which is what really gives the voltage.  This is a variable quantity, as it depends for the given design on rate of rotation.  Sometime there is no simple all encompassing word.  But we do need to try to avoid confusing people.  

« Last Edit: May 24, 2007, 08:47:00 AM by finnsawyer »

bigkahoonaa

  • Jr. Member
  • **
  • Posts: 57
Re: An Alternate PM Alternator Design
« Reply #38 on: July 20, 2007, 09:45:05 AM »
Don't see anything wrong in experimenting with coil design.  I had some extra mag wire and tried something like this:





One massive coil with 6 magnets on either side.  From my limited understanding of magnetic flux, etc., current is only generated in that part of the coil that is perpendicular to both movement of the magnets and direction of the magnetic field (i.e. those parts of the coil going to the center of your stator).  I never tried it with magnets.  I think using a big coil like this one may give starting problems and could cog.


Hope you try your idea some day.

« Last Edit: July 20, 2007, 09:45:05 AM by bigkahoonaa »

finnsawyer

  • Hero Member
  • *****
  • Posts: 1565
Re: An Alternate PM Alternator Design
« Reply #39 on: October 28, 2007, 09:17:35 AM »
There are different ways to look at a situation.  Lately, I've seen a different way of analyzing this alternator design that might help people understand better how Faraday's Law fits it.  As usual, I assume 12 magnets 9 coils for the three phase standard, and identical 12 magnets with identical 18 coils for the design presented here.  When a magnet, say a north pole is directly centered over a coil the flux through the coil is assumed the same for both cases and given the value of F.  In the three phase case as the north pole moves off of the coil and a south pole replaces it the change in flux has a magnitude of 2xF.  This occurs in the time it takes the rotor to move 30 degrees, which I will call t.  So, from Faraday's Law we can write the average voltage V3 induced in the coil as V3 = 2xF/t.


For the design presented here a magnet moves off of or onto a coil resulting in a flux change of magnitude F.  The magnets are spaced every 30 degrees, as before, but the coils are now spaced every 20 degrees.  Furthermore, the geometry is specified to be such that a magnet will move off a coil or onto a coil in a rotation of the rotor of 10 degrees.  So, it takes one third the time for the flux to change by F.  We may then write the average voltage Vm induced in the coil as Vm = F/(t/3) = 3xF/t.  Doing a little trivial manipulation we find that Vm/V3 = 3/2 = 1.5.  That is, Vm = 1.5xV3.

We also recognize that there is essentially no difference in the shape of the pulse of voltage generated when a magnet moves off of a coils as when it moves onto a coil, although the signs differ.  A north pole moving off of a coil will generate a negative pulse, but so will a south moving onto a coil.  Consequently, we can write the total average voltage generated per pulse as 12xVm = 18xV3.


In the case of the three phase alternator we have 3 coils per phase, so the voltage per phase will be 3xV3.  Since two phases almost add in the star configuration, we can write the voltage out from the three phase alternator as 1.73x3xV3 = 5.19xV3.


So, we find that the average voltage out from this proposed design will be 3.47 times as great as from the standard three phase.  What would this mean in practice?  Well, let's assume that cut-in for the three phase occurs at exactly 12 volts due to the battery being low.  That is, the battery voltage plus the diode forward voltage drops equals 12 volts.  Furthermore, the three phase alternator is putting out 18 volts.  Finally, for simplicity the individual coil resistances are 1 ohm.  So, from the three phase we have I3 = (18-12)/6 = 1 amp.  In the case of this design we would have Im = (62.4 - 12)/18 = 2.8 amp, which translates into 2.8 times as much power into the battery.  I've been criticized because of the amount of copper needed, but you see that copper or resistance doesn't tell the whole story.


We have other options.  We could wind the coils two in hand instead of with a single wire.  Doing so cuts the voltage in half, but it also cuts the resistance by a factor of four, with two half as long strands in parallel.  We then would find that we get the following current: Im = (31.2-12)/4.5 = 4.3 amp, even more power into the battery!  Or we could do three in hand for one third the voltage and one ninth the resistance to get: Im = (20.8-12)/2 = 4.4 amp.  Well, there is obviously a limit to this.  Be aware also that you are also affecting cut-in for the alternator by making these changes.  

   

« Last Edit: October 28, 2007, 09:17:35 AM by finnsawyer »

finnsawyer

  • Hero Member
  • *****
  • Posts: 1565
Re: An Alternate PM Alternator Design
« Reply #40 on: October 30, 2007, 08:37:23 AM »
Even though I check the diary from time to time I missed this one.  Sorry about that.  I'm not sure what to make from your statements except to suggest you consider the case of a transformer.  One of the coils takes the place of the magnet and it does not move.  Nevertheless it does create a time varying magnetic flux through the other coil(s).  So, it is the change of flux through the coil with time and not the movement of the magnet per se that is important.  Anyway, your scheme is different than what I propose, as my coils still lie above the magnets in the same orientation as the usual three phase.  That is, the planes of the turns of the coils are parallel to the faces of the magnets.  
« Last Edit: October 30, 2007, 08:37:23 AM by finnsawyer »

pepa

  • Sr. Member
  • ****
  • Posts: 461
Re: An Alternate PM Alternator Design
« Reply #41 on: October 31, 2007, 09:55:03 AM »
Finsawer, email me at suesminisatdotcom, thanks pepa.
« Last Edit: October 31, 2007, 09:55:03 AM by pepa »

pepa

  • Sr. Member
  • ****
  • Posts: 461
Re: An Alternate PM Alternator Design
« Reply #42 on: October 31, 2007, 09:57:58 AM »
at hotmaildotcom,sorry about that, pepa.
« Last Edit: October 31, 2007, 09:57:58 AM by pepa »

tmmjohnson

  • Newbie
  • *
  • Posts: 2
Re: An Alternate PM Alternator Design
« Reply #43 on: November 25, 2007, 10:52:32 AM »
Hi, I am a newbie and just read your posting here.  I am interested in HVLV applications but am finding that it is difficult without just breaking down and buying a unit.  I looked at your idea and thought that maybe I could pass something on that is similar.  It was a college paper done on PMG with toroidial windings.  The purpose was to create a working axial flux generator that could be built in production status with varying scales.  It might be a help, might not be.  I am hoping to get to the point of trying some of the ideas, but magnets are hard to get in Africa. I was trying to get the document copied with its pictures, but I can't seem to paste a picture.  If you give me an address I will send you a PDF of this document.  My address is tmmjohnson@gmail.com.

Jon

NREL/CP-500-24996 Ÿ UC Category: 1213
Axial Flux, Modular, Permanent-
Magnet Generator with a Toroidal
Winding for Wind Turbine
Applications
E. Muljadi
C.P. Butterfield
Yih-Huei Wan
National Wind Technology Center
National Renewable Energy Laboratory
Presented at
IEEE Industry Applications Conference
St. Louis, MO
November 5-8, 1998
National Renewable Energy Laboratory
1617 Cole Boulevard
Golden, Colorado 80401-3393
A national laboratory of the U.S. Department of Energy
Managed by Midwest Research Institute
for the U.S. Department of Energy
under contract No. DE-AC36-83CH10093
Work performed under task number WE803020
July 1998
NOTICE
This report was prepared as an account of work sponsored by an agency of the United States government.
Neither the United States government nor any agency thereof, nor any of their employees, makes any
warranty, express or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents
that its use would not infringe privately owned rights. Reference herein to any specific commercial
product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily
constitute or imply its endorsement, recommendation, or favoring by the United States government or any
agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect
those of the United States government or any agency thereof.
Available to DOE and DOE contractors from:
Office of Scientific and Technical Information (OSTI)
P.O. Box 62
Oak Ridge, TN 37831
Prices available by calling (423) 576-8401
Available to the public from:
National Technical Information Service (NTIS)
U.S. Department of Commerce
5285 Port Royal Road
Springfield, VA 22161
(703) 487-4650
Printed on paper containing at least 50% wastepaper, including 20% postconsumer waste
Axial Flux, Modular, Permanent-Magnet Generator with
a Toroidal Winding for Wind Turbine Applications
E. Muljadi, C. P. Butterfield, Yih-Huei Wan
National Renewable Energy Laboratory
1617 Cole Boulevard
Golden, CO 80401
Tel. (303)384-6900, Fax (303)384-6999
Eduard_muljadi@nrel.gov, http://www.nrel.gov/wind
Abstract - Permanent-magnet generators have been used for
wind turbines for many years. Many small wind turbine
manufacturers use direct-drive permanent-magnet generators.
For wind turbine generators, the design philosophy must
cover the following characteristics: low cost, light weight,
low speed, high torque, and variable speed generation. The
generator is easy to manufacture and the design can be scaled
up for a larger size without major retooling.
A modular permanent-magnet generator with axial flux
direction was chosen. The permanent magnet used is NdFeB
or ferrite magnet with flux guide to focus flux density in the
air gap. Each unit module of the generator may consist of
one, two, or more phases. Each generator can be expanded to
two or more unit modules. Each unit module is built from
simple modular poles. The stator winding is formed like a
torus. Thus, the assembly process is simplified and the
winding insertion in the slot is less tedious.
We built a prototype of one unit module and performed
preliminary tests in our laboratory. Follow up tests will be
conducted in our lab to improve the design.
I. INTRODUCTION
Using permanent-magnet (PM) generators for small wind
turbines is very common. Usually an AC generator with
many poles operates between 10-100 Hz. Because the
generator is directly driven by the wind turbine [1,3,5], it is
commonly known as a direct drive generator. Many
configurations use surface mounted three phase PM
synchronous generators with a rectifier connected to the
generator terminal.
Many types of generator concepts have been used and
proposed to convert wind power into electricity. An axial
flux generator with a different type of winding and a different
magnet arrangement was developed [1,2]. A modular
concept was proposed to reduce manufacturing costs [3]. The
transverse flux generator has a higher power density than a
traditional induction generator [4]. In this paper, a
combination of a modular, axial flux, and torroidal stator
winding are applied to a permanent-magnet generator.
Although the design is intended for wind turbine applications,
this PM machine can be used for many other applications.
A wind turbine generator must be light to minimize the
requirements for the tower structure. Since the wind turbine
operates at low rotational speed, the generator is built with
many poles. We designed, built, and tested a permanentmagnet
generator for wind turbines. Several unique
properties are included in this design. It uses a modular
concept. Each pole is constructed individually, thus the
number of poles is based on the requirements. The winding
is concentric, like a torus, making it easy to assemble. The
rotor core has a focusing capability with a variable magnet
area, so the air gap flux density can be adjusted independent
of the rotor radius. A single unit module of this generator can
have single or multiple phases. Additional unit modules can
be stacked in the axial direction to get more power. With this
modular concept, any failure in one unit can be replaced
immediately or can be bypassed, thus minimizing turbine
downtime.
The dimension of the generator and the size of each
component should be based on the actual wind turbine for
which it is to be used. Because the purpose of the prototype
unit is to prove the concept, we designed and built it with
readily available components. A steady state analysis was
done to determine the initial electric loading and magnetic
loading. The initial loss calculation was derived. The next
step of the calculation was done using finite element analysis.
The flux density in the critical components, and the map of
the core losses were found. No-load, rated, and short-circuit
conditions can be predicted from this analysis. Any changes
made were reiterated by using steady state analysis. Thus the
process was repeated until the final design is ready.
A test was conducted in the lab to find the parameters of
the generator and any unpredicted anomalies. Data were
collected for no-load and full-load conditions.
The first section of this paper is devoted to introducing the
background of the PM generator in wind turbine applications.
The second section introduces the generator components. In
the third section we present our analysis of the PM generator.
In the fourth section we describe testing, and lastly, in the
fifth section the conclusions are summarized.
II. COMPONENT OF THE GENERATOR
In this paper we discuss only one unit module of the
generator. The generator consists of an eighteen-pole
permanent magnet. The stator and the rotor cores are made
of pre-cut transformer lamination silicon steel (gauge 26,
M19). The stator and rotor cores can be made on a per pole
basis, reducing the cost of complete dies required to stamp a
conventional lamination configuration. The geometry of the
stator and the rotor core could have been optimized, however,
this project focuses on the proof of concept.
A. Rotor
The cross section of the stator and rotor pole is illustrated
in Figure 1. Each pole is constructed from two identical corestacks
and the permanent magnet is sandwiched in between.
The rotor is constructed to allow an expansion in the axial
direction, for example, to increase the magnet surface. The
flux directions at the top (outer radius) and the bottom (inner
radius) of the rotor pole are the opposite. Around the
perimeter of the rotor, the flux direction of one pole is
opposite of the flux direction in next pole, as shown by the
white arrows in Figure 2. The ratio of the magnetic surface
area to the pole surface area determines the focusing factor.
The chosen geometry enables the designer to increase the
length of the rotor core without affecting the stator geometry
and vice versa.
The rotor poles are attached to a non-magnetic disk that
holds the rotor cores. The shaft is attached to the disk to
rotate the rotor core. A non-magnetic stainless steel belt is
strapped around the rotor core to keep the rotor poles in
place. Since the rotor speed is low, centrifugal force created
when the rotor rotates is not very high. There are nine pole
pairs on the rotor. Between two rotor poles, there is a small
gap to minimize interpolar magnetic leakage.
B. Stator
The stator consists of two stator sides. There are nine poles
attached to each stator side. The poles on each side are
attached to a plate (not shown in Figure 2) which holds the
stator poles in place. In the prototype, one side of the stator
core can be rotated (within a limited angle range) with respect
to the other stator side. Thus the position of the stator cores
in one side can be shifted with respect to the other sides. The
shift can be adjusted to control the phase shift between the
first stator side and the second stator side.
C. Stator winding
The stator winding is wound like a torus or a washer. With
a toroidal form, the stator winding can be easily assembled
and automated for production. The stator winding between
the stator poles is exposed to open air, which improves
cooling.
One advantage of wind power systems is the location of the
generator. It is mounted on a tower above the ground. The
cooling mechanism is better up on the nacelle than inside a
ground level building because the generator is always
exposed to air flow that is proportional to the generator load.
During low wind speeds, the heat transfer from the winding is
lower, however, the heat generated in the winding is lower,
too. The opposite is true at high wind; more heat is generated
in the winding, but more air flow is available to transfer the
heat away.
In this paper, one module unit is built for a single phase
generator. The stator windings at the two sides are connected
in parallel to generate a single phase output. The rotor shaft
is attached to the stator sides through the bearings, which are
attached to the stator plate. The rotor core has a width of
6.35 cm (2.5 in.) and a diameter of 29.2 cm (11.5 in.). The
overall width of the generator is 16.5 cm (6.5 in.), excluding
the two stator plates.
D. Expansion for multimodule generation system
The power from the stator can be actively controlled using
power switches (IGBTs) or passively controlled using a diode
Figure 1. One pole of the stator and rotor core
Figure 2. PM Generator with Toroidal Winding
stator core rotor core
PM
Non magnet disk rotor core
South
Copper
North
rectifier. Figure 3 shows a possible configuration of the
power converter to process the power generated by the
generator. The generator may consist of one or more
modules. In this configuration, only three unit modules are
shown. Each unit module of the generator is paired with one
leg module of power switches on the power converter side.
Thus the power converter and the generator can be expanded
in a similar fashion. The power generated is converted back
to the utility via a three-phase inverter, which can be
controlled to produce good power quality.
III. DESIGN ANALYSIS
The analysis of the generator is based on the wind turbine
requirements. The steady state analysis was performed as the
first step to get the first cut of design criteria. The finite
element analysis was performed to refine the magnetic
analysis. Finally, a dynamic analysis was performed in the
lab to validate generator performance under dynamic
conditions.
A. Steady state analysis
The prime mover for this generator is a wind turbine. One
characteristic of wind turbines is that the rotational speed is
lower than most prime movers. To avoid using a gearbox,
the generator is direct driven. Multiple poles must be used to
allow slow speed operation.
From steady state analysis, the following criteria are
chosen:
Number of poles = 18
Max operating frequency = 100 Hz (at 667 rpm)
Number of phases per unit module = 1 (two windings in
parallel)
The electric loading:
Stator current = 11.0 Amp RMS (at per phase voltage 58 Volt
RMS)
The wire chosen is AWG 12
The current density in the slot J = 3.4x106 Amp/m2
Predicted copper losses at rated current = 42 watts
B. Finite element analysis
To analyze the magnetic circuit, the finite element method
was used to compute the flux density in the generator
components. The main purpose of this analysis is to get the
overall picture of the saturation levels in different parts of the
generator, the iron losses in the components of the generators,
and the worst case of demagnetization on the permanent
magnet. In the finite element analysis presented here, the
generator uses a ferrite permanent magnet.
No-load condition. In the no-load condition, the magnetic
path is analyzed to see the magnetic flux density in different
parts of the magnetic paths. With the stator core in each side
shifted by 180o the maximum flux in the core happens when
the stator core and the rotor core are aligned. Figure 4 shows
the flux lines at the no-load condition. Only one side of the
stator core is shown. Some flux leakage is shown such as at
both ends of the rotor poles. The rotor core has low flux
density with the highest flux density at the parts closest to the
air gap. As shown in Figure 4, the maximum flux density
Figure 3. Expansion for multimodule generation.
Figure 4. Flux density at no-load condition
Phase3
Phase2
To 60 Hz
utility
Phase1
occurs at the corner of the U-shaped stator core. Figure 5
shows the magnitude of the flux density along the horizontal
line in the middle of the air gap. The maximum flux density
at no load is 1.55 Tesla. The flux density at the air gap is
0.9 Tesla and the flux density at the permanent magnet is
0.24 Tesla. The stator core and the rotor core have a flux
density below the saturation point.
Inductive load at rated current. In this condition, the
magnetic path is analyzed to see flux reduction at the air gap
at the least favorable power factor. The generator is loaded to
have rated current.
Short-circuit condition. In this condition, the magnetic
path is analyzed to see the demagnetization effects on the
permanent magnet. In order to analyze the worst case
scenario, the stator core and the rotor core are perfectly
aligned and the short circuit current is applied to the stator
core. In this case the short circuit current is about ten times
the rated current. The result is tabulated in Table 1.
Table 1. Flux Density Comparison at Different Magnetic
Paths for Different Conditions
B airgap B max B at PM
No-load 0.91 T 1.55 T 0.244 T
Inductive Load (rated) 0.89 T 1.50 T 0.239 T
Short Circuit 0.70 T 1.05 T 0.193 T
IV. EXPERIMENTAL RESULTS
A. Experimental set up
The experiment was conducted to observe the performance
of the generator. The generator is driven by a motor via a
belt. The motor is a four pole motor, with rated speed of
1800 rpm. The motor is fed by a PWM variable frequency
drive. The generator speed is driven to 667 RPM. The
output frequency at this rpm is 100 Hz. The experiment is
conducted only on a single unit generator. In the finite
element analysis, the permanent magnet used is ferrite,
however, in this experiment the permanent magnet chosen is
rare earth permanent magnet (NdFeB).
Figure 6. Open circuit woltage
B. Voltage and current waveforms
The open circuit voltage is measured at the terminal output
of winding 2 (open circuit). The stator cores are shifted
toward each other by 180 electrical degrees. The voltage
waveform is captured from the scope, digitized, and plotted
in Figure 6 and Figure 7.
In Figure 7, the generator is loaded with resistive load up to
rated load at 100 Hz. The voltage across the terminal output
of the generator is a unity power factor load. Thus the
current waveform is reflected by this terminal voltage
waveform.
C. Parameter Determination Test
A simple modified test is used to get the parameters of the
permanent magnet [6]. The experiment is shown in Figure 8.
Figure 5. Flux density at no-load condition
Figure 7. Terminal voltage across resistive load
One side of the generator (winding 1) is connected to a rated
load at unity power factor. The generator is driven to
generate a rated frequency. The other side of the winding
(winding 2) is an open circuit. The voltage output of winding
1 is called terminal voltage V and the open circuit voltage of
winding 2 is called open circuit voltage E. The angle
difference between V and E is called d, which is the torque
angle of the generator at this load. The power, current, and
voltage output of winding 1 is recorded.
The parameters can be computed from the test data, and the
results are listed in Table 2 below.
Table 2. Results from Test Data
Parameters Lds Lqs Rs
8.41 mH 4.38 mH 0.22 ohm
Vopen circuit 75 volts Vrated load 58 volts
Irated/winding 11 Amp Prated/winding 650 watt
Rotor Speed 667 rpm 100 Hz
V. CONCLUSION
The proposed generator is investigated for application in
wind power generation. In the first stage of implementation,
a proof of concept of the generator is investigated. The
magnetic and electric loading are shown to be within the
limits of common practice of machine design. The generator
has the following advantages for wind turbine generation:
- The modular concept is suitable for the commercial
production of machines of limited quantities and with
different sizes and output requirements. The components
are manufactured on a per pole basis. The tooling required
is minimized. The design can be readily changed, such as
the number of poles in one unit or the number of unit
modules in a generator system.
- The axial flux design makes it easier to increase the flux
density in the air gap.
- The toroidal form of the stator winding makes it easy to
fabricate. The geometry of the stator winding and stator
core make the heat dissipation more effective.
- To scale up the output power of the generator, more units
can be stacked in the axial directions. The power converter
required to process the power is readily compatible with
the generator. Each unit module of the generator is
matched with each leg of the power switches.
VI. ACKNOWLEDMENTS
The authors wish to thank Jerry Bianchi for his assistance
during the test set up and Jim Adams for his help during the
fabrication of this generator.
We wish to acknowledge our management at NREL and
the U.S. Department of Energy (DOE) for encouraging us
and approving the time and tools we needed for this project.
DOE supported this work under contract number DE-AC36-
83CH10093.
VII. REFERENCES
[1] B.J. Chalmers, E.Spooner, "An Axial-flux Permanentmagnet
Generator for a Gearless Wind Energy System,"
PEDES 96, January 1996, New Delhi, India.
[2] F. Carrichi, F. Crescimbini, F. Mezzetti, "Multistage
Axial-flux PM Machine for Wheel Direct Drive," IEEE
Transactions on Industry Applications, Vol 32. No. 4,
July/August 1996, pp. 882-887.
[3] E. Spooner, A. Williamson, "Modular, Permanent-magnet
Wind-turbine Generators," Conference Record of the 1996
IEEE Industry Applications Society, Oct. 6-10, 1996, San
Diego, California, Volume 1, pp. 497-502
[4] S. Huang, J. Luo, T.A. Lipo, "Analysis and Evaluation of
the Transverse Flux Circumferential Current Machine,"
Conference Record of the 1997 IEEE Industry Applications
Society, Oct. 5-9, 1997, New Orleans, Louisiana, Volume 1,
pp. 378-384
[5] E.F. Fuchs, A.A. Fardoun, P.Carlin, R.W. Erikson,
"Permanent Magnet Machines with Large Speed Variations,"
Windpower 92, October 1992, Seattle, Washington.
[6] Gieras, J.F., Wing, M., "Permanent Magnet Motor
Technology, Design and Applications," Marcel Dekker, Inc.
New York, 1997.
Figure 8. Experimental set-up
Watt-mtr
/ V,E = d
motor
Winding 1 Winding 2 (open)
V
E
V
Rload
A
V
scope
v
« Last Edit: November 25, 2007, 10:52:32 AM by tmmjohnson »

finnsawyer

  • Hero Member
  • *****
  • Posts: 1565
Re: An Alternate PM Alternator Design
« Reply #44 on: December 03, 2007, 08:27:45 AM »
When I hit stator winding like a torus to make insertion in the slots easier I quit reading, as neither concept would apply to my proposed design.  You really should get a picture.  To post pictures on this site they need to be 640 by 480 pixels and in .jpg format.  One trick you can use to make that conversion is to get a digital camera and take a picture off of the video monitor screen.  If you decide to pursue this any further do an independent posting.  As it is, I do not appreciate all the extraneous verbiage added to the diary, as it simply obscures the real issues.  Another possibility would have been to simply have put a link to the information if such exists.  I would imagine it does.  Rather than simply reacting to things you need to spend more time reading on this site.  Most everything you need to know to navigate on this site is explained somewhere.
« Last Edit: December 03, 2007, 08:27:45 AM by finnsawyer »

1968jeepman

  • Newbie
  • *
  • Posts: 3
Re: An Alternate PM Alternator Design
« Reply #45 on: March 29, 2008, 10:20:02 AM »
I read through all of this , and there have been no additional posts for some time now,  I am wondering if there has been a working model constructed yet?  I am a high school technology education teacher ( shop ).  I am sorry to say, I have little knowledge compared to most posting on this site.  I do however want to build a wind turbine, I have found plans on a magnet/rotor type with a stationary stator.  I to believe  there must be " more " out there.  Other ideas not yet explored.  After reading your posting, I am curious,  like I said before I lack the knowledge, but I can " Build "  I would be very interested and pleased to hear from you directly, if you would be interested to do so.  I have a good deal of free time, and a full shop at my fingertips.  I am hungry to learn all I can.  Thanks for your time.
« Last Edit: March 29, 2008, 10:20:02 AM by 1968jeepman »

TomG

  • Jr. Member
  • **
  • Posts: 70
Re: An Alternate PM Alternator Design
« Reply #46 on: June 07, 2008, 09:49:36 PM »
Absolutely fascinating! It's now 4:50am where I am, and I've spent the past several hours analysing this, and its implications.


If I understand it correctly, your alt essentially has 3 different cutin points.


Below the first cutin, obviously nothing happens.

Between the first and second cutins, the EMF is higher than 12V for 1/3rd of the time.

Between the second and third cutins, the EMF is higher than 12V for 2/3rd of the time.

After the third cutin, the EMF is higher than 12V all the time.


This leads to a non-linear increase in the current flow, as you showed in:

http://www.otherpower.com/images/scimages/918/POW1.jpg


Very clever! The alt gradually steps up its duty cycle (0% -> 33% -> 66% -> 100%) as the RPM increases!


Actually, this happens to some extent in a standard alt, too. Because the waveform generated by the coils is sinusoidal (or triangular, but certainly not square), when the peak EMF first passes 12V, only the very peak of the waveform is above 12V. Current only flows during that time, so the duty cycle is low. As the peak EMF increases, the width of the section of waveform above 12V rapidly increases, and the duty cycle approaches 100%.


If the waveform is a sine wave:


It's a bit better if the waveform is a triangle wave, because the tip is pointier, so the cutin is more gradual:


Your alt's waveform is a 3-step approximation of a triangle wave (if you move the sections about a bit - their order makes no difference).


All this led me to come up with a design of my own:

http://www.fieldlines.com/story/2008/6/8/34644/48584

« Last Edit: June 07, 2008, 09:49:36 PM by TomG »

finnsawyer

  • Hero Member
  • *****
  • Posts: 1565
Re: An Alternate PM Alternator Design
« Reply #47 on: June 10, 2008, 09:24:39 AM »
This entire post was intended to illustrate the principle by which the design would work.  That's why the square pulses.  Of course, with actual coils having finite width windings that would not happen.  Unfortunately, that seemed to confuse a number of people.  Well, you do the best you can.


The actual pulse shape will also depend on when the peaks occur in the two cases of a magnet moving onto a coil or off a coil.  In the 12 magnet 18 coil version there will be 6 of each such actions happening for each half cycle.  So the voltages add in pairs to give the final shape.  Interestingly, that shape should be the same as for a single phase from the three phase alternator if one makes the time scales the same.  It's not obvious for the three phase, as it involves two magnets and one coil, but the principle of superposition says it should be the same.


Be curious to see how this "breakpoint" idea finally works out for you.

« Last Edit: June 10, 2008, 09:24:39 AM by finnsawyer »

TomG

  • Jr. Member
  • **
  • Posts: 70
Re: An Alternate PM Alternator Design
« Reply #48 on: June 10, 2008, 05:06:17 PM »
This entire post was intended to illustrate the principle by which the design would work.  That's why the square pulses.  Of course, with actual coils having finite width windings that would not happen.  Unfortunately, that seemed to confuse a number of people.


No, I got that that was just for illustration. But my point still holds, even if the waveform of each coil is roughly a sine wave (or triangle wave). The sum waveform of the one-and-a-half phases which gives the breakpoint/gradual cutin effect still approximates a triangle wave.


The actual pulse shape will also depend on when the peaks occur in the two cases of a magnet moving onto a coil or off a coil.  In the 12 magnet 18 coil version there will be 6 of each such actions happening for each half cycle.  So the voltages add in pairs to give the final shape.  Interestingly, that shape should be the same as for a single phase from the three phase alternator if one makes the time scales the same.  It's not obvious for the three phase, as it involves two magnets and one coil, but the principle of superposition says it should be the same.


Hmm. I think I've managed to work out what that waveform should be, but do you have a sketch of what you'd expect it to look like?

« Last Edit: June 10, 2008, 05:06:17 PM by TomG »

finnsawyer

  • Hero Member
  • *****
  • Posts: 1565
Re: An Alternate PM Alternator Design
« Reply #49 on: June 11, 2008, 08:22:17 AM »
I don't know about the triangular aspect, but considering that leakage flux and other factors may come into play one really needs to look at what happens in practice.  As I mentioned, there is somewhere on this site a posting that shows the actual pulse shapes on a scope picture for the single coil, single magnet interaction, but I couldn't find it, and it has not surfaced since.
« Last Edit: June 11, 2008, 08:22:17 AM by finnsawyer »

TomG

  • Jr. Member
  • **
  • Posts: 70
Re: An Alternate PM Alternator Design
« Reply #50 on: June 11, 2008, 12:13:47 PM »
True, a prototype would be nice, but I'm not going to be able to build one any time soon.


In lieu of that, an explanation of what I was talking about with the gradual cutins.


If the waveform going to a rectifier is a square wave, the cutin will hard. A sudden change from no current (0% duty cycle) to some current always flowing (100% duty cycle).

If the waveform going to a rectifier is a sine wave, the cutin will be fairly abrupt, because the waveform has a fairly flat top to it. We can see this more clearly if we "sort" the waveform, bring this portions with similar values together. This is physically unrealistic, but would actually make very little difference to the output of the rectifier if we did it in real life (ideally, no difference).


If the waveform going to a rectifier is a triangle wave, the cutin will be more gradual, because the waveform has no flat top to it. It gradually eases above the battery voltage, as the coil EMF rises.



(The steps are an artifact of my program only analysing 360 discrete steps. IRL it would be a smoother, nearly straight line)


The waveform from one-and-a-half phases of your design, (which is what has the interesting waveform and the breakpoints in it) approximates the results from a triangle wave:



The three plateaus which cause the three distinct breakpoints can be clearly seen when it's sorted like that.

« Last Edit: June 11, 2008, 12:13:47 PM by TomG »