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 The information provided may be out of date.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.

Modelling coils near resonance frequency

Please login with a confirmed email address before reporting spam

Hi,

I'm modelling an eddy current testing probe using the AC/DC module. It essentially consists of a transmitter coil with an applied input current and also a receiver coil which will have an induced voltage. I'm modelling both coils in 3D using the multi-turn coil domain.

I've got the actual probe and I've been testing the model experimentally, but so far I haven't been able to get the model to agree with experiment. The reason seems to be that there is resonance in the receiver coil near the frequencies that I'm interested in, but COMSOL isn't taking that into account. Does anyone know if there is a way to model the coils so that I can find the induced voltage across the receiver near the resonance frequency? Or are there any useful tutorial models I could follow?

Thanks,

Neil

5 Replies Last Post 2013年7月18日 GMT-4 16:59
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年7月17日 GMT-4 16:32
Neil,

modeling a real coil is not easy. The multiturn coil domain doesn't account for winding capacity and thus doesn't exhibit self-resonance as a real coil does.

You could model the coil explicitly, i.e. model each winding including winding separation, insulation, etc. This way winding capacity would be accounted for, but of course, the mesh size would probably explode.

Or you could try to connect the coil to an external circuit, which would then essentially be a parallel capacitor to mimic the self resonance.

Cheers
Edgar

--
Edgar J. Kaiser
www.emphys.com
Neil, modeling a real coil is not easy. The multiturn coil domain doesn't account for winding capacity and thus doesn't exhibit self-resonance as a real coil does. You could model the coil explicitly, i.e. model each winding including winding separation, insulation, etc. This way winding capacity would be accounted for, but of course, the mesh size would probably explode. Or you could try to connect the coil to an external circuit, which would then essentially be a parallel capacitor to mimic the self resonance. Cheers Edgar -- Edgar J. Kaiser http://www.emphys.com

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年7月18日 GMT-4 13:26
Hi Neil,

As was noted in the first response above me, modeling coils near resonance isn't easy. In AC/DC, in fact, it really can't be done explicitly.

However, going to the RF module will allow for taking into account not just the coil inductance (as the AC/DC module does), but it also captures the coil self-capacitance. Again, capacitance isn't captured in the AC/DC module explicitly.

The RF module has a tutorial example called "RF coil" or something like that and it shows how to do this type of modeling using both the eigenvalue solver to get the resonant frequency, and then the frequency domain sweep to get the Q-factor of the coil.

It also shows how to do it relatively inexpensively in terms of computation resources by neglecting the skin depth (either using impedance BC on the coils or perfect electriic conductor...the only difference being that the former accounts for coil losses).

Maybe I didn't say anything new as compared to the response here already, but thought I'd add about the RF coil example.

Good luck!

--Mattt
Hi Neil, As was noted in the first response above me, modeling coils near resonance isn't easy. In AC/DC, in fact, it really can't be done explicitly. However, going to the RF module will allow for taking into account not just the coil inductance (as the AC/DC module does), but it also captures the coil self-capacitance. Again, capacitance isn't captured in the AC/DC module explicitly. The RF module has a tutorial example called "RF coil" or something like that and it shows how to do this type of modeling using both the eigenvalue solver to get the resonant frequency, and then the frequency domain sweep to get the Q-factor of the coil. It also shows how to do it relatively inexpensively in terms of computation resources by neglecting the skin depth (either using impedance BC on the coils or perfect electriic conductor...the only difference being that the former accounts for coil losses). Maybe I didn't say anything new as compared to the response here already, but thought I'd add about the RF coil example. Good luck! --Mattt

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年7月18日 GMT-4 16:27

Hi Neil,

As was noted in the first response above me, modeling coils near resonance isn't easy. In AC/DC, in fact, it really can't be done explicitly.

However, going to the RF module will allow for taking into account not just the coil inductance (as the AC/DC module does), but it also captures the coil self-capacitance. Again, capacitance isn't captured in the AC/DC module explicitly.

The RF module has a tutorial example called "RF coil" or something like that and it shows how to do this type of modeling using both the eigenvalue solver to get the resonant frequency, and then the frequency domain sweep to get the Q-factor of the coil.

It also shows how to do it relatively inexpensively in terms of computation resources by neglecting the skin depth (either using impedance BC on the coils or perfect electriic conductor...the only difference being that the former accounts for coil losses).

Maybe I didn't say anything new as compared to the response here already, but thought I'd add about the RF coil example.

Good luck!

--Mattt


Thanks, but unfortunately my coil has 19,300 turns so modelling it explicitly isn't really an option. (And the RF module
can't do any of that without modelling it explicitly, can it?)
[QUOTE] Hi Neil, As was noted in the first response above me, modeling coils near resonance isn't easy. In AC/DC, in fact, it really can't be done explicitly. However, going to the RF module will allow for taking into account not just the coil inductance (as the AC/DC module does), but it also captures the coil self-capacitance. Again, capacitance isn't captured in the AC/DC module explicitly. The RF module has a tutorial example called "RF coil" or something like that and it shows how to do this type of modeling using both the eigenvalue solver to get the resonant frequency, and then the frequency domain sweep to get the Q-factor of the coil. It also shows how to do it relatively inexpensively in terms of computation resources by neglecting the skin depth (either using impedance BC on the coils or perfect electriic conductor...the only difference being that the former accounts for coil losses). Maybe I didn't say anything new as compared to the response here already, but thought I'd add about the RF coil example. Good luck! --Mattt [/QUOTE] Thanks, but unfortunately my coil has 19,300 turns so modelling it explicitly isn't really an option. (And the RF module can't do any of that without modelling it explicitly, can it?)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年7月18日 GMT-4 16:32

Neil,

modeling a real coil is not easy. The multiturn coil domain doesn't account for winding capacity and thus doesn't exhibit self-resonance as a real coil does.

You could model the coil explicitly, i.e. model each winding including winding separation, insulation, etc. This way winding capacity would be accounted for, but of course, the mesh size would probably explode.

Or you could try to connect the coil to an external circuit, which would then essentially be a parallel capacitor to mimic the self resonance.

Cheers
Edgar

--
Edgar J. Kaiser
www.emphys.com


Thanks Edgar, your second idea sounds good. Do you mean just measuring the capacitance experimentally and adding it to the circuit? I will try that.
[QUOTE] Neil, modeling a real coil is not easy. The multiturn coil domain doesn't account for winding capacity and thus doesn't exhibit self-resonance as a real coil does. You could model the coil explicitly, i.e. model each winding including winding separation, insulation, etc. This way winding capacity would be accounted for, but of course, the mesh size would probably explode. Or you could try to connect the coil to an external circuit, which would then essentially be a parallel capacitor to mimic the self resonance. Cheers Edgar -- Edgar J. Kaiser http://www.emphys.com [/QUOTE] Thanks Edgar, your second idea sounds good. Do you mean just measuring the capacitance experimentally and adding it to the circuit? I will try that.

Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年7月18日 GMT-4 16:59

Neil,

yes that's what I would suggest. In a first attempt you could just add a parallel capacitor value resulting from the analytical formula for LC resonance and see if the results meet your expectations.

Good luck
Edgar

--
Edgar J. Kaiser
www.emphys.com
Neil, yes that's what I would suggest. In a first attempt you could just add a parallel capacitor value resulting from the analytical formula for LC resonance and see if the results meet your expectations. Good luck Edgar -- Edgar J. Kaiser http://www.emphys.com

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.