[Getdp] Eddy Current Testing in getDP

[Frederic Trillaud]

Dear Judith,

I am actually trying to compute induced current in a superconductor and
have struggled not with the A-V formulation but I am trying out T-Omega
formulation.
As far as I understood you are solving two equations both having A and V
but lumped one into the other, should you solve the equations in a
coupled manner. Saying this, I am not an expert in electromagnetism and
slowly entering the field so any comments may be thoroughly checked!!

In first approximation if the magnetic field is vertical in the sheet of
metal or can be concidered as close to it (along the axis of symmetry),
you can assume V = 0 and solve: Curl nu0 Curl A+sig*dA/Dt-Jsource = 0
and compute the induced current in the sheet as sig*dA/dt and see if it
is reasonable and from there you build on it.

Sorry not be of more help, best,

Frederic




-----Original Message-----
From: Mengelkamp Judith TU Ilmenau <Judith.Mengelkamp at tu-ilmenau.de>
To: getdp at geuz.org <getdp at geuz.org>
Subject: [Getdp] Eddy Current Testing in getDP
Date: Fri, 19 Jun 2015 12:18:37 +0000

Dear getDP-users,

 

I am currently struggling with simulating Eddy Current Testing in getDP.
I have implemented a simplified model (one turn as an excitation coil
with sinusoidal source current and a metal sheet as test specimen). I
use an A-V-formulation (magnetic vector potential and electric scalar
potential). I use a theta-time scheme for the time-varying source
current in the coil. My current .geo and .pro-files are in the appendix.
The magnetic field of the coil is calculated correctly in time, however
the eddy current induced in the conductor result into zero and I don’t
have an explanation for that. Further, the magnetic vector potential
doesn’t change in time and the electric potential is not defined in all
volume elements of the air. I would be very pleased if someone could
help me with these issues. Here are some remarks concerning my
implementation:

-         The fifth integral term in the Formulation constraints the
gradient of the electric potential at the surface of the conductor.
Because I didn’t know how to implement this as a constraint, I put this
term in the formulation and use a user-defined function for the normal
component. Is this ok?

-         I set the electric potential in a randomly chosen point to
zero to ensure the solution is unique.

-         I also noticed that the magnetic field in the conductor is
relatively high and does not change in time.

 

Because I am quite new to getDP...is there any paper or similar that
describes finite element calculations with eddy currents?

 

Thank you very much in advance,

Judith.

 

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

M.Sc. Judith Mengelkamp

 

Technische Universität Ilmenau

Institut für Biomedizinische Technik und Informatik (Institute for
Biomedical Engineering and Informatics)

Fachgebiet Biomedizinische Technik (Biomedical Engineering Group)

 

Phone: +49 3677 69-1958

Fax:       +49 3677 69-1311

e-mail:  judith.mengelkamp at tu-ilmenau.de

 


_______________________________________________
getdp mailing list
getdp at geuz.org
http://www.geuz.org/mailman/listinfo/getdp