COMSOL Forums: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/
Most recent forum messagesMon, 04 Feb 2013 05:14:22 +0000COMSOL Forums: sinusoidal voltage to time dependent solverhttp://www.comsol.com/shared/images/logos/comsol_logo.gif
http://www.comsol.com/community/forums/general/thread/33436/
Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p95083
Hi<br />
<br />
the phasor approach (complex amplitudes), should work check the the doc<br />
<br />
--<br />
Good luck<br />
IvarMon, 04 Feb 2013 05:14:22 +000033436.1359954862.95083Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p95055
<div class="quote"><br />
Hi<br />
<br />
have you got your model running ?<br />
At least you could reduce the overall air size to some diam 0.4 ad somewhat longer than you imported model, then add some 4 cm of infinite elements and set them up for cylindrical coordinates. But I'm not sure what you expect to see ?<br />
<br />
then you define the ima,b,c over all the domains, but they have some sense only locally, I would rather use the formula directly in the multi-Turn Coil BCs<br />
<br />
Anyhow for continuous AC you should use the frequency sweep mode at a fixed frequency of 60 Hz and then give the amplitudes as complex values (physors) for the phase instead of the wv1,2,3(t) approach. The time domain is only useful for a transient turn on, and in MF you do not get any ringing as the 2nd time derivative physics is ignored<br />
<br />
<br />
--<br />
Good luck<br />
Ivar<br />
</div><br />
<br />
thanks. I think I have to give up using time dependent analysis with this model. But the problem is phase shift. How can I consider Ib=120<-240deg (i.e. 120 amplitude and -240degree of phase angle). I can define this in time dependent but I don't know how to define it in frequency sweep. please help<br />
<br />
Regards,Sun, 03 Feb 2013 21:46:33 +000033436.1359927993.95055Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p94719
Hi<br />
<br />
have you got your model running ?<br />
At least you could reduce the overall air size to some diam 0.4 ad somewhat longer than you imported model, then add some 4 cm of infinite elements and set them up for cylindrical coordinates. But I'm not sure what you expect to see ?<br />
<br />
then you define the ima,b,c over all the domains, but they have some sense only locally, I would rather use the formula directly in the multi-Turn Coil BCs<br />
<br />
Anyhow for continuous AC you should use the frequency sweep mode at a fixed frequency of 60 Hz and then give the amplitudes as complex values (physors) for the phase instead of the wv1,2,3(t) approach. The time domain is only useful for a transient turn on, and in MF you do not get any ringing as the 2nd time derivative physics is ignored<br />
<br />
<br />
--<br />
Good luck<br />
IvarMon, 28 Jan 2013 19:42:23 +000033436.1359402143.94719Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p92856
Hi fellas<br />
<br />
I finally could have time-dependent analysis with magnetic field solution. however it has a serious problem. I used MF solver with multi turn coil for each phase of my machine. the problem is working for just a single phase and when I enable the boundary for the second and/or third phase, the solver doesn't work. I tried many modification but no progress. I would appreciate if you would help me. the file is attached (the solution is removed because of the large size: 800MB)<br />
<br />
Regards, <br />
Sun, 23 Dec 2012 23:09:02 +000033436.1356304142.92856Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91917
Hi<br />
<br />
if you are studying a steady state sinus, it often far easier and quicker to use the frequency domain solver than a true transient, transient behaviour and sustained sinus excitation is not fully the same, because you have all the transient effects from your inperfect BC's in the time solver<br />
<br />
--<br />
Good luck<br />
IvarFri, 07 Dec 2012 15:55:11 +000033436.1354895711.91917Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91914
what a comprehensive answer. thanks.<br />
<br />
you mentioned " you run time solving in EC and a parametric sweep in MF based on the current domain mapping" . would you mention the details or refer to a document about this coupling?<br />
<br />
Regards,Fri, 07 Dec 2012 14:55:48 +000033436.1354892148.91914Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91819
Hi<br />
<br />
check <a href="http://www.comsol.eu/community/forums/general/thread/33429/#p91817" title="www.comsol.eu/community/forums/general/thread/33429/#p91817">www.comsol.eu/community/forums/general/thread/33429/#p91817</a> and see if that can inspire you ;)<br />
<br />
--<br />
Good luck<br />
IvarThu, 06 Dec 2012 19:40:35 +000033436.1354822835.91819Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91811
thanks for the answer. the final question from is this: is there anyway to apply voltage and see magnetic field while using time dependent solver?<br />
<br />
Thanks for support<br />
<br />
Thu, 06 Dec 2012 18:41:55 +000033436.1354819315.91811Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91803
Hi<br />
<br />
there is in the doc if you get to the right page ;), but you should also check the solver proposed when you enter the physics, there you see which are "set-up" by default, and also by looking to the equations , if you have d/dt 's your time response will work if you have w omega's, the frequency response too, but if none appear, when you switch your selected solvers then no luck.<br />
<br />
Also ACDC is assuming instantaneous time response on the magnet fields, else you are in wave propagation mode = RF<br />
<br />
--<br />
Good luck<br />
IvarThu, 06 Dec 2012 15:30:37 +000033436.1354807837.91803Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91800
great answer. thank you very much. I don't know why there is no note about incompatibility of MEF with time dependent solver in manuals. Also thanks for your note. but one question rises now<br />
<br />
if magnetic and electric interface doesn't work for this, how can we see magnetic field while we apply voltage to the model. another question is " Is it OK to use parametric solver with the time as variable instead of time dependent one with magnetic and electric solver?"<br />
<br />
Regards,<br />
Thu, 06 Dec 2012 14:43:06 +000033436.1354804986.91800Re: sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91789
Hello,<br />
<br />
The Magnetic and Electric fields physics does not support Time-dependent studies, only Stationary and Frequency. You can see it in the Time Dependent step (the yellow sign near the physics's name in the table).<br />
<br />
Depending on what you want to do, you need to use another physics, for example Electric Currents.<br />
Note that your model as it is now cannot be solved with a Magnetic interface since it does not respect current conservation (current is "created" on one boundary and "destroyed" on another).<br />
<br />
--<br />
Andrea Ferrario<br />
Electromagnetics Group<br />
COMSOL ABThu, 06 Dec 2012 09:17:21 +000033436.1354785441.91789sinusoidal voltage to time dependent solver
http://www.comsol.com/community/forums/general/thread/33436/#p91738
I am trying to apply sinusoidal voltage to a model with time dependent solver. I apply the voltage based on a function but it doesn't show anything. the model is attached. <br />
<br />
thanksWed, 05 Dec 2012 20:40:13 +000033436.1354740013.91738