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 The information provided may be out of date.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.
Transient simulation output doesn't match input
Posted 29 nov. 2012, 05:39 UTC−5 Low-Frequency Electromagnetics, Studies & Solvers Version 4.3a 5 Replies
Please login with a confirmed email address before reporting spam
Hi
I'm simulating a 2mm radius magnetic loop excited by a transient current and surrounded by a 10cm air sphere, in time dependent study.
I've interpolated a gaussian current input from several points along time from 0 to 12E-8 (attached file).I'm exciting the loop using magnetic field (mf)'s external current density node along phi (cylindrical system) with "1[A]*int1(t[1/s]))/0.00000019635[m^2]". Note that I've declared "t" as a variable in seconds.
After running the time dependent simulation in "range(0,1.0e-8,1.2e-7)", I wanted to plot the magnetic field of a cut point located at 50mm (x,y and z) along time using the 1D point graph which, basically, must result in a gaussian curve too.
The problem is that I get a very different plot (not gaussian at all) with weird magnetic field values. Another problem is that the curve of the magnetic field with z component (attached) is different from the magnetic field, y component (attached) and also the x component, which I think should have been more or less the same according to my researches.
I tested with a sinusoidal excitation and it's the same case.
So, could you please tell me what's wrong with my simulation. Please help, it's means a lot to to me .
Thanks very much.
I'm simulating a 2mm radius magnetic loop excited by a transient current and surrounded by a 10cm air sphere, in time dependent study.
I've interpolated a gaussian current input from several points along time from 0 to 12E-8 (attached file).I'm exciting the loop using magnetic field (mf)'s external current density node along phi (cylindrical system) with "1[A]*int1(t[1/s]))/0.00000019635[m^2]". Note that I've declared "t" as a variable in seconds.
After running the time dependent simulation in "range(0,1.0e-8,1.2e-7)", I wanted to plot the magnetic field of a cut point located at 50mm (x,y and z) along time using the 1D point graph which, basically, must result in a gaussian curve too.
The problem is that I get a very different plot (not gaussian at all) with weird magnetic field values. Another problem is that the curve of the magnetic field with z component (attached) is different from the magnetic field, y component (attached) and also the x component, which I think should have been more or less the same according to my researches.
I tested with a sinusoidal excitation and it's the same case.
So, could you please tell me what's wrong with my simulation. Please help, it's means a lot to to me .
Thanks very much.
Attachments:
5 Replies Last Post 9 janv. 2013, 06:28 UTC−5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I used a Waveform excitation instead of the interpolation. I declared a sinusoidal current of 2MHz frequency and I excited the coil using "(1[A]*wv1(t[1/s]))/0.000019635[m^2]". Attached file.
I run the time dependent simulation in the range "0, 10-7, 50 10-7". Still, the simulation output of the magnetic field doesn't match the output, even if it's closer this time, since I don't find a sinusoidal plot. Attached also.
What's causing the results to be this inacurrate, please ? Any indication could help me.
Thanks
I used a Waveform excitation instead of the interpolation. I declared a sinusoidal current of 2MHz frequency and I excited the coil using "(1[A]*wv1(t[1/s]))/0.000019635[m^2]". Attached file.
I run the time dependent simulation in the range "0, 10-7, 50 10-7". Still, the simulation output of the magnetic field doesn't match the output, even if it's closer this time, since I don't find a sinusoidal plot. Attached also.
What's causing the results to be this inacurrate, please ? Any indication could help me.
Thanks
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
are you sure the physics is still valid for 2MHz and your model size ? ACDC assumes wavelength much greater than the model size, such that magnetic effects are "instantaneous w.r.t. the model. Normally the harmonic development and frequency domain solver is better suited for sinus type excitations
--
Good luck
Ivar
are you sure the physics is still valid for 2MHz and your model size ? ACDC assumes wavelength much greater than the model size, such that magnetic effects are "instantaneous w.r.t. the model. Normally the harmonic development and frequency domain solver is better suited for sinus type excitations
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thanks very much for the reply Ivar. You're really the angel of Comsol :)
It seems you're right. I tested with a 2 Hz only excitation wave and it reasulted in a correct sinusoidal output. Like you said, it seems the wave's frequency in mega hertz is so high that comsol is assuming it's instantenous. So does that means that the magnetic field (mf) physics and all of ACDC just can't work for this excitation and that there's no hope nor solution ?
Am I obliged in this case to go for Radio frenquency physics, knowing that I'm searching to study electromagnetic field in transient domain ? To tell the truth I worked so hard to get this running in ACDC that I really prefer to find any solution there, instead of returning from the start.
Is there any solution please ?
Thanks very much
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
one ends often up in the intermediate region where ACDC and RF are both valid or only partially, often one should make both models and compare. Check carefully the hypothesis of both physics
--
Good luck
Ivar
one ends often up in the intermediate region where ACDC and RF are both valid or only partially, often one should make both models and compare. Check carefully the hypothesis of both physics
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
I decided to use the RF module, particularly the Electromagnetic Waves, transient (temw). I used a "surface current" node for the excitation and I specified "1[A]*wv1(t[1/s]))/0.015708[m]" [A/m] along phi inside the dipole, with the 0.015708 is 2*pi*dipole's radius. And a perfect magnetic conductor node for the air surronding the dipole.
The wv1 is a sinusoidal wave of 2MHz. And I simulated using time dependant study setting "range(0,5.0e-11,5.0e-7)"
Still, even with the RF module, the simulation output doesn't match the input since the field in fonction of the time isn't sinusoidal (attached). Even if, with this physic too, a wave of 50Hz does give an output corresponding to the input.
Is it a limitation from Comsol concernng the high frequencies or am I missing something ? Is there any solver configuration or any way to resolve this problem? Please help ?
The wv1 is a sinusoidal wave of 2MHz. And I simulated using time dependant study setting "range(0,5.0e-11,5.0e-7)"
Still, even with the RF module, the simulation output doesn't match the input since the field in fonction of the time isn't sinusoidal (attached). Even if, with this physic too, a wave of 50Hz does give an output corresponding to the input.
Is it a limitation from Comsol concernng the high frequencies or am I missing something ? Is there any solver configuration or any way to resolve this problem? Please help ?
Attachments: