Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Model BH-curves with remanent flux density at H=0
Posted 25 feb 2014, 05:37 GMT-5 Low-Frequency Electromagnetics, Materials 4 Replies
Please login with a confirmed email address before reporting spam
Hello,
I make some studies to determine the magnetic properties of metals and have to take in account a small remanent magntic flux density B.
I defined a sample material with a known BH-curve and at H=0 I set a small value for B to set the remanent flux density..
But I always get the error no solution found even with minimum step size.
Is there a other possibilits to simulate nonlinear materials with a small remanent flux density?
Thanks for your advices.
I make some studies to determine the magnetic properties of metals and have to take in account a small remanent magntic flux density B.
I defined a sample material with a known BH-curve and at H=0 I set a small value for B to set the remanent flux density..
But I always get the error no solution found even with minimum step size.
Is there a other possibilits to simulate nonlinear materials with a small remanent flux density?
Thanks for your advices.
4 Replies Last Post 28 feb 2014, 04:56 GMT-5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hallo Florian,
you have to remember that H and B are really vector quantities. If you simply shift
the BH curve there is no direction defined. I assume that you use
Hnorm = f(Bnorm) + Bremenace
But in which direction is the remance pointing. I guess that is why you
system is not well defined and Comsol is complaining
Regards
Jens
you have to remember that H and B are really vector quantities. If you simply shift
the BH curve there is no direction defined. I assume that you use
Hnorm = f(Bnorm) + Bremenace
But in which direction is the remance pointing. I guess that is why you
system is not well defined and Comsol is complaining
Regards
Jens
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
@Jens
Thanks for the advice,
I have done it as you said (Hnorm = f(Bnorm) + Bremenace) and I pointed out that this is the problem,but I do not find another possibility to combine nonlinear material and remanent flux density for a given direction.
Regards
Florian
Thanks for the advice,
I have done it as you said (Hnorm = f(Bnorm) + Bremenace) and I pointed out that this is the problem,but I do not find another possibility to combine nonlinear material and remanent flux density for a given direction.
Regards
Florian
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
One idea:
in Ampere's law -> Magnetic field you can change the constitutive relation to Magnetisation.
The you can enter a vector. e.g.
Mx = M0x + fx(Bx,By,Bz)
My = fy(Bx,By,Bz)
Mz = fz(Bx,By,Bz)
which would be for permeant magnetisation in x-direction. The functions fx,fy,fz rerpesent
a vectorial M(B) relation. For isotropic material this could be
fx=F(normB)/normB*Bx and y,z accordingly
F now is a function with F(0)=0 representing a scalar M(B) law..
A bit of a work to implement, but should do the job.
Good luck,
Jens
in Ampere's law -> Magnetic field you can change the constitutive relation to Magnetisation.
The you can enter a vector. e.g.
Mx = M0x + fx(Bx,By,Bz)
My = fy(Bx,By,Bz)
Mz = fz(Bx,By,Bz)
which would be for permeant magnetisation in x-direction. The functions fx,fy,fz rerpesent
a vectorial M(B) relation. For isotropic material this could be
fx=F(normB)/normB*Bx and y,z accordingly
F now is a function with F(0)=0 representing a scalar M(B) law..
A bit of a work to implement, but should do the job.
Good luck,
Jens
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
@Jens
Thanks this is a good idea with your advice I could solve it nearly the same way.
Solution:
-I had a BH-curve for my material and converted it into a curve murB (mur = f(B)).
-In the ampere's Law node for the material I chose "Remanent Flux Density (B =mu0*mur*H+Br)"
-For relative permeability I chose "User defined" and "Isotropic".
- The function I inserted was "f(mf.normB[1/T])". The function must depend on B , if it depends on H you have a circulary dependency error.
-The remanent flux density was set to some value in x-direction.
Now I think it is working.
Thanks for the help and maybe my solution can be helpful for others.
Florian
Thanks this is a good idea with your advice I could solve it nearly the same way.
Solution:
-I had a BH-curve for my material and converted it into a curve murB (mur = f(B)).
-In the ampere's Law node for the material I chose "Remanent Flux Density (B =mu0*mur*H+Br)"
-For relative permeability I chose "User defined" and "Isotropic".
- The function I inserted was "f(mf.normB[1/T])". The function must depend on B , if it depends on H you have a circulary dependency error.
-The remanent flux density was set to some value in x-direction.
Now I think it is working.
Thanks for the help and maybe my solution can be helpful for others.
Florian