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.
Modelling AC currents in 2D axi
Posted Mar 5, 2010, 10:31 a.m. EST 9 Replies
Please login with a confirmed email address before reporting spam
Hello;
i would like to model a circuit (hollow cylinder for example) with current flowing in the r-z plan and not in the azimutal direction (phi) which is the default setting in Comsol multiphysics. (The magnetic field would be along phi and not in the plane).
I thought that i could write the PDE equation and adapt the ampere law with no current in the phi direction.
But, the PDE formulation is not available in 2D axi. why? (is it because it is only available in the AC/DC module?)
Best regards
benjamin
i would like to model a circuit (hollow cylinder for example) with current flowing in the r-z plan and not in the azimutal direction (phi) which is the default setting in Comsol multiphysics. (The magnetic field would be along phi and not in the plane).
I thought that i could write the PDE equation and adapt the ampere law with no current in the phi direction.
But, the PDE formulation is not available in 2D axi. why? (is it because it is only available in the AC/DC module?)
Best regards
benjamin
9 Replies Last Post Mar 8, 2010, 10:54 a.m. EST
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
Indeed the PDE is not there by default, I believe that it is because to get the 2D-axi to run flawless, you need to add some mathematical subtilities such as a *r to avoid the singularity on axis, which means that postprocessing need to take the /r to get the correct results. probably this means that it is not evident to mix in your own PDEs.
But you can still make your own 2D axi with a 2D mode and redefine the variables and the equations, while beng inspired by what COMSOL has done in their 2D-Axis.
Check the variables, it would be an excellent exercice to learn the internals, if you have th time.
Good luck
Ivar
Indeed the PDE is not there by default, I believe that it is because to get the 2D-axi to run flawless, you need to add some mathematical subtilities such as a *r to avoid the singularity on axis, which means that postprocessing need to take the /r to get the correct results. probably this means that it is not evident to mix in your own PDEs.
But you can still make your own 2D axi with a 2D mode and redefine the variables and the equations, while beng inspired by what COMSOL has done in their 2D-Axis.
Check the variables, it would be an excellent exercice to learn the internals, if you have th time.
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
thanks a lot, but do you know if with the AC/DC module in 2D axi you can choose the current direction (phi or r-z plan?)
Best regards
benjamin
thanks a lot, but do you know if with the AC/DC module in 2D axi you can choose the current direction (phi or r-z plan?)
Best regards
benjamin
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I'm not 100% sure, but I have not found how to define a current differently than radially, but how would you represent the current in 2D axi if not as function of phi ?
Ivar
I'm not 100% sure, but I have not found how to define a current differently than radially, but how would you represent the current in 2D axi if not as function of phi ?
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
the current could be defined as a function of r. The structure that i try to model is an axi symetric hollow conical horn. There is a current flowing in the inner and outer conductor. This produces a toroidal magnetic field. (B is in, out the plan only function of r and z). In DC mode, we can defined an inward current in the r-z plan.
Regards
benjamin
the current could be defined as a function of r. The structure that i try to model is an axi symetric hollow conical horn. There is a current flowing in the inner and outer conductor. This produces a toroidal magnetic field. (B is in, out the plan only function of r and z). In DC mode, we can defined an inward current in the r-z plan.
Regards
benjamin
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
Could it be that I have left the plasma physics for too long time ago ?
but I cannot visualise your shape just from the wording, is it possible to have a sketch view ?
Just f0r fun ?
Ivar
Could it be that I have left the plasma physics for too long time ago ?
but I cannot visualise your shape just from the wording, is it possible to have a sketch view ?
Just f0r fun ?
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
i have seen in the AC/DC book that we can choose between the 2D in plane currents application modes and meridional currents mode with the currents present in the rz plane. That's exaclty what i need!
regards
benjamin
i have seen in the AC/DC book that we can choose between the 2D in plane currents application modes and meridional currents mode with the currents present in the rz plane. That's exaclty what i need!
regards
benjamin
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
i have attached the geometry of the electromagnetic horn. In this horn, they will be a strong magnetic field between the conductor. The peak current is 300kA! the voltage is 7000V, the material aluminum probably.
Benjamin
i have attached the geometry of the electromagnetic horn. In this horn, they will be a strong magnetic field between the conductor. The peak current is 300kA! the voltage is 7000V, the material aluminum probably.
Benjamin
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
In my undrstanding, the meridional or toroidal current selection are usefull for solving the magnetic field, but in your cse it's simply Joules heating, no? then its less critical this should work correctly for 2D axi.
I would also go for the "segregated solver" as you problem can be split, and for the moment you have not linked the temperature to the material properties.
But with these powers you will end up vaporising your system (in steady state) if you do not cool it better.
Finally take care when workg in 2D axis,: you have to get the units to match, you need to often multiply by "2*pi*r" the "loop length" for units coherence.
This should be done explicitely in your formulas for the integration variables, but can be selected with a "click" for the postprocessing.
Always carefully check the units to be sure you havnt forgotten the 2*pi*r !!
And do your homework with simple analytical formulas to be sure you have the correct orders of magnitude for your results
Good luck
Ivar
In my undrstanding, the meridional or toroidal current selection are usefull for solving the magnetic field, but in your cse it's simply Joules heating, no? then its less critical this should work correctly for 2D axi.
I would also go for the "segregated solver" as you problem can be split, and for the moment you have not linked the temperature to the material properties.
But with these powers you will end up vaporising your system (in steady state) if you do not cool it better.
Finally take care when workg in 2D axis,: you have to get the units to match, you need to often multiply by "2*pi*r" the "loop length" for units coherence.
This should be done explicitely in your formulas for the integration variables, but can be selected with a "click" for the postprocessing.
Always carefully check the units to be sure you havnt forgotten the 2*pi*r !!
And do your homework with simple analytical formulas to be sure you have the correct orders of magnitude for your results
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
in fact, the current will be sinusoidal (50Hz) with a pluse of 100us of 300kA everry 20ms, so most of the current is flowing in the skin depth. We also want to have the magnetic field distribution; (we already have with analytical calculation).
We will a water cooling system for the walls. I also plan to do thermal stress analysis and modal analysis.
regards
ben
in fact, the current will be sinusoidal (50Hz) with a pluse of 100us of 300kA everry 20ms, so most of the current is flowing in the skin depth. We also want to have the magnetic field distribution; (we already have with analytical calculation).
We will a water cooling system for the walls. I also plan to do thermal stress analysis and modal analysis.
regards
ben