Long Simulation Time for Time-Dependent Solver

Please login with a confirmed email address before reporting spam

Hi,

I am modeling an electromagnet with DC input at a constant voltage with the aim to understand the time-dependent changes in the conductivity, flux density, temperature and pull force of the equipment.

The target cycle time is 0-24 hours with steps of 1 hour. The core is soft iron with a CRGO block domain at its center.

I am unable to attach the .mph file since it is currently still running.

The time-solver is progressing time in "seconds" and not "hours" and the simulation for every 5 seconds is taking 30 seconds in real time.

The same applies for studies without CRGO that I have done for 0-8 hour results which have spanned for 12-14 hours in real time.

I need all the results mentioned before, but I need to understand how I can bring down the simulation time drastically since the real time consumed is too long.

Thank you.



2 Replies Last Post 31.05.2025, 18:37 MESZ
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 4 days ago 31.05.2025, 06:21 MESZ

If you haven't done it already, consider specifying the time step size manually, to a value appropriate to the physics you are modeling. Default time steps are often not the ideal choices. In most cases, you know (or should know) more about the (approximate) rates of change of important quantities than the code can guess. It is good software, but it is not an AI system -- you can provide the code with the wise guidance it needs!

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
If you haven't done it already, consider specifying the time step size *manually*, to a value appropriate to the physics you are modeling. Default time steps are often *not* the ideal choices. In most cases, you know (or should know) more about the (approximate) rates of change of important quantities than the code can guess. It is good software, but it is not an AI system -- you can provide the code with the wise guidance it needs!

Please login with a confirmed email address before reporting spam

Posted: 4 days ago 31.05.2025, 18:37 MESZ

Sometimes Fully Coupled is much faster. But it depends. Posting a model, even a model with a simplified geometry, may lead to some useful suggestions.

Sometimes Fully Coupled is much faster. But it depends. Posting a model, even a model with a simplified geometry, may lead to some useful suggestions.

Reply

Please read the discussion forum rules before posting.

Please log in to post a reply.

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.