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.

Magnetic Field Simulation; 'Stationary' and 'time dependent study' simultaneously

Please login with a confirmed email address before reporting spam

This is Jason Kim, studying biomedical engineering in Korea in the field of diagnostic medical device. I'd like to discuss and share your experiences regarding 'Coil and magnetic field' and here I post the question here. The link below is my precedented question and answer.

https://www.comsol.com/forum/thread/224241/stationary-and-time-dependent-study-at-the-same-time?last=2018-12-21T09:49:35Z

Basically, I am trying to find an magnetic interaction between 'rotating permanent magnet' in the magnetic field which is genereated by 'a pair of multi-turn coil' [Pic 1].

By the way, I had a problem with computing both 'time-dependent' and 'stationary study' at together. Below are what I tried but failed.

Independently computing 'time-dependent' and 'stationary study' one by one. ->Each result was established without error but not interacted. The magnetic field from magnet and coil is not mixed but independtly existed. [Pic 2, Pic 3]

Processing all physics only with Time-dependent study ->Error: Failed to finding Initial vaue and Internal error [Pic 4]

Right click 'study(time-dependent)' and tried 'Combined solution' ->Error: Cannot concatenate different types[Pic 5]

I think there must be a solution to solve this problem and I beg you a insight for arranging these studies. Thank you.

Jason



2 Replies Last Post 2 gen 2019, 10:42 GMT-5
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 6 years ago 26 dic 2018, 05:04 GMT-5

Jason,

I am not sure if I fully understand your objective, but may be following hints can be useful.

For a time dependent problem it can be useful to do a stationary study step and then take the stationary solution as an initial value for the time dependent step. But you have to take care that the time dependent situation at t=0 approximates the stationary situation. E.g. in a rotating system calculating stationary without rotation and then start the rotation at full speed may not work. You need to ramp up the speed from zero. Just an example, your problem may be different.

You cannot combine a stationary and a time dependent solution. As said before stationary can be a good initial value for time dependent.

Cheers Edgar

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Jason, I am not sure if I fully understand your objective, but may be following hints can be useful. For a time dependent problem it can be useful to do a stationary study step and then take the stationary solution as an initial value for the time dependent step. But you have to take care that the time dependent situation at t=0 approximates the stationary situation. E.g. in a rotating system calculating stationary without rotation and then start the rotation at full speed may not work. You need to ramp up the speed from zero. Just an example, your problem may be different. You cannot combine a stationary and a time dependent solution. As said before stationary can be a good initial value for time dependent. Cheers Edgar

Michael Rembe Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 6 years ago 2 gen 2019, 10:42 GMT-5
Updated: 6 years ago 3 gen 2019, 04:15 GMT-5

Hi Jason, hi Edgar,

At first I wish you all the best for 2019!!

I think it is possible to combine a stationary solution with a transient solution within a for-loop.

For this you have to divide your transient problem into (nmax-1) time periods and you repeat a for loop (-> solution -> programming) nmax times.

At first (before the loop starts) you have to compile the quations and to set the initial values (Dependent variables). Then the for loop starts. The work flow is as follows:

  • compile equations
  • dependent variables (set initial values)
  • For (begin)
    • compile equations
    • dependent variables (set initial values to current)
    • instationary solver
    • compile equations
    • dependent variables (set initial values to current)
    • stationary solver
  • end for

What we need is the for-loop counter to recalculate the time (start, end) for the transient solver. This is currentiter, using this varable you can recalculate the correct time.

You have to take care, that the transient solution runs using the stationary solution.

If you need some additional help, please contact me.

Cheers, Michael Rembe

-------------------
Michael Rembe
Rembe Consulting PartG mbB
www.rembe-consulting.de


Hi Jason, hi Edgar, At first I wish you all the best for 2019!! I think it is possible to combine a stationary solution with a transient solution within a for-loop. For this you have to divide your transient problem into (nmax-1) time periods and you repeat a for loop (-> solution -> programming) nmax times. At first (before the loop starts) you have to compile the quations and to set the initial values (Dependent variables). Then the for loop starts. The work flow is as follows: * compile equations * dependent variables (set initial values) * For (begin) * compile equations * dependent variables (set initial values to current) * instationary solver * compile equations * dependent variables (set initial values to current) * stationary solver * end for What we need is the for-loop counter to recalculate the time (start, end) for the transient solver. This is currentiter, using this varable you can recalculate the correct time. You have to take care, that the transient solution runs using the stationary solution. If you need some additional help, please contact me. Cheers, Michael Rembe

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.