[femm] Re: Various

[David Meeker <dmeeker@xxxxxxxxxxxxxxxxx>]

Agnaldo Souza Pereira wrote:

>Does anybody know what viewer may I use for opening the
>"mesh maker" program called "triangle"?

What are you trying to do? Look at the executable itself? It doesn't seem
like that would be very enlightening. If you want to see how the program does
what it does, the full source to triangle is available at:
http://www.cs.cmu.edu/~quake/triangle.html

--------------------------
Agnaldo Souza Pereira wrote:

>> I have data sheets on some of the materials, but on the majority,
>> I just have the BH information.
>
> Hi, David! These BH curves can be very useful to me.
> Could you send them to me?

All of the info that I have as far as the virgin magnetization curves of
various materials is already imbedded in the materials library in femm. You
can open up the materials library, go to the material of interest,and open up
the BH curve to get the information. The BH data can be cut-and-pasted out of
femm to an editor or something.

--------------------------
Dave Squires wrote:

> David,
>
> I think I found a bug in your code.
> I was trying to change the Hc on the same model
> from one run to the next and it was not being reflected
> in the results. It seems you do not reload the
> materials data for each run. I have to delete it from
> the model list, run to make it fail and put it back it
> to force a reload. Also, modifying parameters from
> the "materials" button does not copy them into the
> library.
>
> To avoid such a problem I suggest a slight change
> to the code so that it will always reload all materials
> data for the problem when the "solve" button is pushed.
> This would guarantee any changes in the model library
> would be reflected in the new run.

This isn't a bug--it just works a little bit differently than the way that you
are thinking about it. The materials library is meant to be just a repository
for materials that are commonly used. You shouldn't modify library entries
themselves, unless you are entering a new material into the library or
something. When you are in the Materials Library and press the "Add selected
material to model" button, it makes a copy of the library entry and puts it in
the list of materials properties being used your specific model. To change the
properties employed in solving the model, you change the properties of the
_copy_ of the library entry, via the Properties|Materials selection off of the
main menu. The reason for doing it this way is that I don't want it to mess
around with library entries (which I will presumably use over and over in
future models) each time I want to change the properties for a specific
problem.

--------------------------
Victor Bellido-Gonzalez wrote:

> Does anyone know of any comprehensive help or tutorial on GMSH and
> GetDP, especially regarding 3D magnetostatic modelling?
> Any basic designs or models that could help me starting on it?

I don't think that these programs are very well documented in general. Your
best bet is to get in contact with the authors--they might have some better
documentation of samples to offer that is not available on the web. A link to
their website (well, you probably already know it if you have the program, but
anyway...) is: http://www.montefiore.ulg.ac.be/services/elap/elm_en.html

--------------------------
riad kechrpud wrote:

> I am riad. I am from Algeria. I am preparing a PhD
> thesis on the modelization of induction motors taking
> into account the movement Could help me?. I am looking
> for the magnetisation curve B(H) of XC18. Can the femm
> be adapted to take into account the movement?

femm doesn't really do problems with motion in the way that you'd really like
for the analysis of induction motors. I have femm to look at induction motors
with motion in the past, but this is somewhat tricky. I've set up two
problems--one is just the stator with an inner boundary half-way across the air
gap. The other is just the rotor with an outer boundary half-way across the
air gap. By doing multiple runs with different harmonic distributions of A on
the boundaries, it is possible to get sort of impedance relationships between
the different harmonics--I'm basically using the model to identify a
complicated but parametric model of the machine that includes higher
harmonics. As I say, however, this is all sort of tricky and really only works
well for linear materials.

It would not be impossible to modify the program to include motion, but it
would take more time than I am likely to have in the near future. I'm planning
to release the 3.0 version that I'm working on under the Aladdin Free Public
License (sort of like Gnu Public License, but a stronger in terms of
restricting people from charging money for modifying and/or redistributing the
code), so the source will eventually be available if you are interested in
modifying it. However, since you are working on your PhD, you'd get more out
of it if you wrote your own program that is just specific to the machines that
you are interested in looking at. It's not as hard as it sounds, particularly
if it is a program with only a command-line user interface. Doing the user
interface and coding things in a general way so as to be applicable to a large
number of problems was what really took the time in writing femm.
Comparatively, the real numerical coding was quick.

Dave.



begin:vcard 
n:Meeker;David
tel;fax:781-890-3489
tel;work:781-684-4070
x-mozilla-html:TRUE
url:http://members.aol.com/dcm3c
org:Foster-Miller, Inc.;Electrical and Electronic Systems Group
version:2.1
email;internet:dmeeker@xxxxxxxxxxxxxxxxx
title:Senior Engineer
adr;quoted-printable:;;350 Second Avenue=0D=0A;Waltham;MA;02451-1196;USA
fn:David Meeker, Ph.D.
end:vcard