Hi,
With the ACDC module of Comsol, I would like to calculate the standard parameters of air coil by taking into account frequency effect (inductance, resistance and capacitance of the coil)
I would like to run this Comsol model under Matlab in order to make a parametrical optimization. This mean that the solve procedure of the model should not be too long !
I already read the model library of ACDC module, but still didn't find the best solution for my model.
Lumped parameters can be used to do that. But as I understand, I need to use a 3D model and apply potiential (GND, port) to the coil terminal (the "power inductor" model of the model library or the "planar transformer" used such Lumped parameters to calculate the Impedance of the coil.
I tried to modelize a 1 layer 30 turns coils in 3D by using the helix3 command line in Matlab (remember that the geometrical parameters of the coil must be tunable under matlab for a future optimization procedure), then apply potential (GND and port) to the coil terminal as in power inductor example. Unfortunately the solver doesn't converge. I'm a newbie in Comsol so I don't know which parameters to modify in the solver (I keep the same as in the power inductor example).
BUT this way to do seems to take a lot of time to solve.
For sure, it is also possible to make an approximation by using a one "area-equivalent" conductor in a 2D axis-symetrical environnement by definiing a current density in the conductor and then using a Psi/I expression to calculate the inductance of the coil (I also have to calculate mutual inductance between 2 coils). I also to that in the static-magnetic mode... but this doesn't take into account frequency effect.
which is the best way to compute inductance or mutual inductance simulation in 2D ?
I also wanted to take into account the resistance increase functions of the frequency, this can be done eventually by modeling 1 conductor per layer and then multiplying the resistance by the number of turns per layer in the coil...
Simulation of the whole 3D coil seems to be a crazy way, isnt'it ?
Thanks for the help!
With the ACDC module of Comsol, I would like to calculate the standard parameters of air coil by taking into account frequency effect (inductance, resistance and capacitance of the coil)
I would like to run this Comsol model under Matlab in order to make a parametrical optimization. This mean that the solve procedure of the model should not be too long !
I already read the model library of ACDC module, but still didn't find the best solution for my model.
Lumped parameters can be used to do that. But as I understand, I need to use a 3D model and apply potiential (GND, port) to the coil terminal (the "power inductor" model of the model library or the "planar transformer" used such Lumped parameters to calculate the Impedance of the coil.
I tried to modelize a 1 layer 30 turns coils in 3D by using the helix3 command line in Matlab (remember that the geometrical parameters of the coil must be tunable under matlab for a future optimization procedure), then apply potential (GND and port) to the coil terminal as in power inductor example. Unfortunately the solver doesn't converge. I'm a newbie in Comsol so I don't know which parameters to modify in the solver (I keep the same as in the power inductor example).
BUT this way to do seems to take a lot of time to solve.
For sure, it is also possible to make an approximation by using a one "area-equivalent" conductor in a 2D axis-symetrical environnement by definiing a current density in the conductor and then using a Psi/I expression to calculate the inductance of the coil (I also have to calculate mutual inductance between 2 coils). I also to that in the static-magnetic mode... but this doesn't take into account frequency effect.
which is the best way to compute inductance or mutual inductance simulation in 2D ?
I also wanted to take into account the resistance increase functions of the frequency, this can be done eventually by modeling 1 conductor per layer and then multiplying the resistance by the number of turns per layer in the coil...
Simulation of the whole 3D coil seems to be a crazy way, isnt'it ?
Thanks for the help!