Can someone please tell me where the sticky point is, and maybe how to reduce it?
The rotating parts are iron, and the parts inside the outer stator magnets are also iron.
Quote from: Low-Q on April 05, 2007, 12:07:17 PM
Can someone please tell me where the sticky point is, and maybe how to reduce it?
The rotating parts are iron, and the parts inside the outer stator magnets are also iron.
Hi,
very nice motor !
Can you please post the FEMM files in each position , so I can have a look at the calculated torque
output from the rotor ?
You can calculate the torque with FEMM over the whole cycle and then see, if there are
more positive torque components or if all torque components will
sum up to zero...
But this looks like there is always a positive torque or
at least most of the time !
That?s very good !
Regards, Stefan.
Quote from: hartiberlin on April 05, 2007, 08:00:22 PM
Quote from: Low-Q on April 05, 2007, 12:07:17 PM
Can someone please tell me where the sticky point is, and maybe how to reduce it?
The rotating parts are iron, and the parts inside the outer stator magnets are also iron.
Hi,
very nice motor !
Can you please post the FEMM files in each position , so I can have a look at the calculated torque
output from the rotor ?
You can calculate the torque with FEMM over the whole cycle and then see, if there are
more positive torque components or if all torque components will
sum up to zero...
But this looks like there is always a positive torque or
at least most of the time !
That?s very good !
Regards, Stefan.
I haven't stored all positions, but I have used -10 degrees for each position. All lines in the rotating part is "curve". and the center position is 0, 0.
Idon't know how to measure torque.
Hi Vidar,
Nice animation!
I can't use femm quite well, but if i can recall correctly, you can measure torque by clicking on parts / blocks of your simulation, then click on the integrate button and choose torque from the list.
Here is 18 saved FEM-files, -10 degree for each file. I cannot find the integrate button (How does it look like?), not in the file menu...
Br.
Vidar
Okay, here is attached to this message the first torque measurement.
YOu must just draw an additional circle around the object inside a smaller
triangled airgap and close the circle with the red circle
drawing tool in the output Blux density diagram and then click the
the Line integral symbol and click Torque from Stress tensor.
This gives about 1.2 NewtonMeter torque at the zero degrees rotation setting.
I will now see, what it is for each single rotor.
Regards, Stefan.
Forgive my ignorance, but would the motor just stop about the position in the below image?
Quote from: hartiberlin on April 09, 2007, 07:24:14 PM
Okay, here is attached to this message the first torque measurement.
YOu must just draw an additional circle around the object inside a smaller
triangled airgap and close the circle with the red circle
drawing tool in the output Blux density diagram and then click the
the Line integral symbol and click Torque from Stress tensor.
This gives about 1.2 NewtonMeter torque at the zero degrees rotation setting.
I will now see, what it is for each single rotor.
Regards, Stefan.
Thanks for your help!
I look forward to se the final result, even if it is negative :)
BTW: You must probably exclude the fixed center magnet in your measurements to get a correct torque value (?)
Br.
Vidar
I finally made it, Stefan. I excluded the inner permanent magnet, and got -5,5Nm torque instead of -1,2Nm.
I also closed the iron ring and got -0,04Nm torque...I believe this is average torque in one revolution.
I'll try anoter angles on the inner permanent magnet to se if it alters the torque.
Br.
Vidar
Hi Vidar,
yes, the inner circle magnet must be excluded.
I saw, that it gives greatly different results, if you include it !
I must redraw all your drawings first to make
more space for another line integral line around the moving
rotors and also make the airgap triangles mesh more dense at this
point as this lowers the errors.
Also it helps in faster computing to define the pure iron material
as Linear and not Nonlinear, which accelerates the computation a lot
if you have a finer mesh.
The introduced error is not so big with this easier calculations.
Sorry, I am pretty busy with other things right now and I will
try to get to it during this week.
Stay tuned.
You have to make a graph of all the torque values versus rotor
angles and see, if the positive values over a 360 degree cycles are
more than then negative torque values over this cycle.
If it does not add up to ZERO, then we have a positive result
and the motor will selfrun, if the simulation is not wrong.
If I only had time to learn this LUA Script language, all would be much easier..
Regards, Stefan.
Quote from: bitRAKE on April 09, 2007, 07:34:35 PM
Forgive my ignorance, but would the motor just stop about the position in the below image?
If you would attach a flywheel to the rotor axis these angles, where there is
not much positive torque could be overcome, if the integral over the torque
over a complete 360 degrees cycle is still positive !
Then we would get a selfrunning motor.
If the integral of the torque over the complete cycle adds up to zero, then it would not work.
Regards, Stefan.
I have done some modifications to the motor by using three iron poles instead of two. I have also added a steel can to the outer magnets. In addition, there is now fully magnet on the outer magnets - no iron parts attached to them. See the picture for details.
I have now measured an average of 9 Nm in 360degrees. All measurements had a torque in the same direction (?!).
So I was thinking: Isn't it more correct to use the "Area" instead of "Contur", tool and mark the rotating parts alone to measure torque in the rotating iron parts - as all the magnets are fixed as stators?
If I do that, the torque during 360 degrees are not so promising...
Br.
Vidar
Hi Vidar,
yes the 3 single rotors must be measured alone and added up.
I think the contur line will give the same as the Area tool.
it only depends on the meshes that are next to the rotors inside the airgap,
there are ocurring the forces.
So put a "trapez-circle" line inside the airgap around each rotor and then measure the torque
from stress tensor for each rotor and add this up for each rotor angle.
If you know how to use LUA scripts, you could do this in 1 degrees rotor angle intervals
and then get every sum of the rotor torques and put in into Excel or Calc
spreadsheet from OpenOffice and plot a torque versus rotor angle graph.
Then are there more positive rotor torques more than negative rotor torques
values ?
Add up their areas and you can see, if it adds up to zero or
to a positive or negative value.
If it is zero then it will not work, if it is unequal to zero, then the motor will run
either clockwise or counterclockwise.
Regards, Stefan.
Hi Vidar,
I have used FEMME extensively and have written script files that will auto-rotate and take the torque measurements. If you attach that latest FEMME file, I'll be happy to run the simulation for you and generate a torque plot.
God Bless,
Jason O
yes, Jason,
if you could please do this, this would be great,
as I don?t have time in this moment to play with it and
I would need to do it all manually, cause I don?t know LUA script language yet.
Many thanks in advance.
Regards, Stefan.
Hello All,
here is a torque analysis done on the first simulation file posted bu Vidar. I increased the mesh size to 0.3 for all block labels and did a block torque analysis only on the two magnets rather than a line integral around the whole rotating part. There are 72 torque points plotted, meaning the measurement was taken every five degrees around the circle for the total 360 degree rotation.
There is definitely a clear offset on the graph though I wonder if the total area still integrates to zero. I'll be happy to see how the one with three rotors works.
God Bless,
Jason O
Hi Jason,
many thanks for this fantastic fast result.
Could you do some kind of script calculation
over one cycle and see if the area under the curve
in the positive range is the same area as under the zero line
in the negative range ?
If both areas are the same , the torque adds up to zero and
the motor will not run.
It could be that it really adds up to zero from your posted graph.
So we would need to tweak the design, so that it does not add up
to zero.
This could probably be done by overlapping rotors or
added flux diverser iron pieces inside the stator or rotor, so the negative torque would
not be this big in some places...
Regards, Stefan.
Hi Stefan,
Attached is the Excel file I use to generate the torque graphs. I input the torque data files from the simulation and it will automatically format the output for me. There are also a few tables on there that show the total positive torque points, the total negative torque points, and an average of the two. This is just an approximation of the area of the graph to show if it is more positive or negative.
It is showing that there is slightly more positive (0.7%) but since the ratio is so small, I'm willing to bet that there is an equal area around the graph. But, even with that, it is the negative spike that I am interested in. Because of what I talked about in my 90 degree rule article (www.fdp.nu/free_energy.asp?book=90), it is the spike that we want to bring in the extra energy into the system. Check out the article for more info.
Also keep in mind that the simulator cannot account for the many complex magnetic effects which could allow this to work. So I wouldn't get discourages just because it doesn't give positive results.
God Bless,
Jason O
I believe the simplest way to see the total torque, is to replace the two or three rotor parts with one single toroid - a ring shaped iron piece surrounding the inner permanent magnet. Then you have in theory an infinite number of degrees to calculate the average torque. If the magnetic field through the iron ring, in any way, will force the ring to rotate one way, one can probably make a working model.
Unfortunately the femm file for my last three pole motor does not longer exist. Femm has the great feature which don't ask you to save the file before closing if you have done changes to the previous saved model. So now this motor is suddenly a total different model without asking me first... ::)
Br.
Vidar
what software is being used to produce the images and torque numbers.
Hi d3adp00l,
I created the torque data files using a Lua script I made for FEMME. The graphs are made inside a Microsoft Excell spreadsheet.
God Bless,
Jason O
not to sound to dumb but what is femme?
Hi d3adp00l,
FEMME (Finite Element Method Magnetics) is a freeware electromagnetic simulator that Vidar and I used to do magnetic simulations with. You can read more about and download it here:
http://femm.foster-miller.net/wiki/HomePage
God Bless,
Jason O
Not FEMME (French: girl) but FEMM ;)
Anyway, Femm 4.1 can be downloaded for free at this site:
http://femm.foster-miller.net/wiki/Download (http://femm.foster-miller.net/wiki/Download)
Br.
Vidar
Well I can make little rectangles but thats about it, its a bit over my head to use. But thanks for lettin me know where it was.
Quote from: d3adp00l on April 15, 2007, 12:13:28 AM
Well I can make little rectangles but thats about it, its a bit over my head to use. But thanks for lettin me know where it was.
Look at the tutorial how to do it. I did and after just a few minutes, and a few questions here, I can manage to use it, measure torque, force etc.
You first have to make a square or a circle where the device will be inside. Use the point tool to click four points where you then use the line tool to draw a line between the points. You are now creating a space where the device can work in.
Br.
Vidar
I will watch the tutorial and take notes. Thanks, I have the start, I am kinda excited after about ten years waiting to figure out how to create the field I needed I get try and test it.