[femm] Re: Hi! (for Agnaldo)

[frederic.sirois@xxxxxxxxxx]

Dear Agnaldo,

I just discovered this workgroup today. I was very surprised to see
that you
are trying to model superconductors, because it is rather uncommun. I
do have
exactly the same interest as you, though not exactly the same
applications in
view. More precisely, the modelling of superconductors using finite
elements
for applications in power engineering is the aim of my PhD thesis, in
Montreal,
Canada. I have tried a commercial package called "Magnet 5", from
Infolytica corporation, a Montreal based compagny. It was not a great
success. Now I'm investigating the possibility of writing my own FE
package to model the material properties. Since it is intended to be an
engineering optimization tool, my approach is to use field dependant
power law resistivity curves to model the superconductor (see J.
Rhyner, "Magnetic properties and AC-losses of superconductors with a
power law current-voltage characteristics," Physica C 205 (1993)
292-300.), with a possibility of varying the characteristics across the
sample to make it inhomogeneous. That way, it would be possible to
account for strong surface pinning and Meissner screening, though it is
not totally formaly correct. As a first approach, it sounds simpler
that solving the Ginzburg-Landau equations.

I will certainly have a look soon at your paper. I'm curious about how
you modelled both the Meissner shielding and the mixed state using an
unmodified commercial package.

Frederic


agnaldo souza pereira <agnald-@xxxxxxxxxx> wrote: 
original article:http://www.egroups.com/group/femm/?start=65
> Dear David,
> my main interests are related the utilization of superconductors in
magnetic
> bearings. My group has developed a methodology for simulate the total
and
> partial magnetic shielding due to meissner and mixed states, using a
> commercial FEM program, without the needing of modifying that program
or
> writing a new one(at the same time a spanish group has presented the
same
> methodology).
> We have used this method to calculate the levitation forces in a
magnetic
> bearing that was built in the electrical engineering graduate program
at
> COPPE-UFRJ. Our theoretical result for the levitation height was 3.8
mm and
> the experimental result measured by the engineers was about 3.5 mm.
> Now I think we are gonna solve the Ginzburg-Landau equations of
> superconductivity, by FEM, in order to construct the B(H) curve for a
> superconductor. We have already done this curve with the method
mentioned
> above, but using the Bean model.
> Another problem we'd like to deal with is the simulation of the
magnetic
> response of a superconductor to a time-varying magnetic field.
> I'd like to discuss some theoretical aspects of FEM and would be very
> pleased in helping to incorporate a superconductor calculation
capability
> into femm.
> 
> For a better understanding of superconductivity I suggest you to read
the
> book "Introduction to Superconductivity" by Rose-Innes and Rhoderick,
and of
> course we could talk about the subject here in the list, what could
bring
> benefits to other people who are interested in this subject.
> I'm sending you (in attachment)a resumed form of our work. 
> An article of ours is in IEEE Transactions on Applied
Superconductivity, 
> june 1999,vol.9 number 2, page 964.
> 
> Agnaldo.
> 
> 
> 
> >>> Agnaldo
> >
> >In founding this mailing list, I was hoping that we'd get some of
this sort 
> >of discussion going. That is, understanding the sort of problems
that are 
> >important to other people, and hopefully learning something new in
the 
> >process.
> >
> >Agnaldo--I am aware of a few of the "odd" properties of
superconductors, but 
> >I don't understand them thoroughly enough to include them in femm. 
What 
> >behaviors are you trying to model? Would you be interested in
helping to 
> >incorporate a better understanding of superconductors into femm? 
(This goes 
> >for everyone--it might be interesting to get more manpower working
on making 
> >improvements and extensions to the program.)
> >
> >Cristi--Similar questions. The finite element analysis of
electromagnets 
> >covers a lot of territory. What particular applications are you
interested 
> >in, and what sorts of extensions do you envision to present
techinques to 
> >address the problems that you are looking at?
> >
> >Dave.
> >--
> >http://members.aol.com/dcm3c
> >