Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Induction heating

Please login with a confirmed email address before reporting spam

Hi to all!
I'm getting some trouble about induction heating using frequency-transient solver.
In the 2D-axisymmetric model, I have an aluminium workpiece surrounded by 12 turn of coil (each one setted in "single turn coil domain 1").
In the "single turn coil domain 1" I set the power of the coil, and here is the problem! If i set very low power (10-60 W) the simulation works so good, but if I set the power I need (at least 6000 W) I get the error
"Time-Dependent Solver 1 (sol1/t1)
Failed to find consistent initial values"
I think I have to change some BC, but i don't know which one!

Anyone knows?

Thank you!

6 Replies Last Post 5 mag 2013, 05:38 GMT-4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mag 2013, 06:16 GMT-4
Hi

perhaps it's simply that your initial conditions are too far from the solution so jumping straight to 6kA is too stiff, have you tried to use a step function and let COMSOL ramp up the current from nearly 0 to your max value with a smmoth step or ramp ?, in 3-10 steps ?

--
Good luck
Ivar
Hi perhaps it's simply that your initial conditions are too far from the solution so jumping straight to 6kA is too stiff, have you tried to use a step function and let COMSOL ramp up the current from nearly 0 to your max value with a smmoth step or ramp ?, in 3-10 steps ? -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mag 2013, 06:30 GMT-4
Hi Ivar, and thanks for answering!
So, the problem might be about to coil, passing from 0 W to 6000 W in a smallest time step?

PS: why you talked about current and not power?
Hi Ivar, and thanks for answering! So, the problem might be about to coil, passing from 0 W to 6000 W in a smallest time step? PS: why you talked about current and not power?

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mag 2013, 07:04 GMT-4
Hi

just a glimps, I usually increase current in coils, but you are right it can also be power ;) Whatever drives your model. Do not forget that mesh, time steps and gradients are often linked by some fundamental formulas, so too steep gradients with finite mesh gives sometimes issues for the solver convergence. For that the parametric sweep are useful to help convergence, and to follow what is happening (use the plot while solving ...)

--
Good luck
Ivar
Hi just a glimps, I usually increase current in coils, but you are right it can also be power ;) Whatever drives your model. Do not forget that mesh, time steps and gradients are often linked by some fundamental formulas, so too steep gradients with finite mesh gives sometimes issues for the solver convergence. For that the parametric sweep are useful to help convergence, and to follow what is happening (use the plot while solving ...) -- Good luck Ivar

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mag 2013, 07:04 GMT-4
Hi

just a glimps, I usually increase current in coils, but you are right it can also be power ;) Whatever drives your model. Do not forget that mesh, time steps and gradients are often linked by some fundamental formulas, so too steep gradients with finite mesh gives sometimes issues for the solver convergence. For that the parametric sweep are useful to help convergence, and to follow what is happening (use the plot while solving ...)

--
Good luck
Ivar
Hi just a glimps, I usually increase current in coils, but you are right it can also be power ;) Whatever drives your model. Do not forget that mesh, time steps and gradients are often linked by some fundamental formulas, so too steep gradients with finite mesh gives sometimes issues for the solver convergence. For that the parametric sweep are useful to help convergence, and to follow what is happening (use the plot while solving ...) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mag 2013, 09:20 GMT-4
Thanks again Ivar for you support!
I read that setting power as input induces nonlinear equations, so could be better to set current.
But i would set both voltage and current, because if i set just one of them i don't know which value has the other one.
How i can do it? If i set current, there is a setting that allow me to set voltage too?
Thanks again Ivar for you support! I read that setting power as input induces nonlinear equations, so could be better to set current. But i would set both voltage and current, because if i set just one of them i don't know which value has the other one. How i can do it? If i set current, there is a setting that allow me to set voltage too?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 5 mag 2013, 05:38 GMT-4
any ideas?
any ideas?

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.