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.
PDE solver with non-linear terms
Posted 30 gen 2025, 15:31 CET Equation-Based Modeling Version 6.3 1 Reply
Please login with a confirmed email address before reporting spam
Hi COMSOL Community,
I am trying to solve a PDE that I derived, which does not follow the conventional equation formats available in the existing modules. Therefore, I am using the PDE solver in COMSOL. To define my governing equation, I had to use two coefficient PDEs (u1 and u2) and link them together.
I am also utilizing the Generalized Alpha Implicit solver; however, my simulation fails after a few time steps when using an initial condition. I am unsure whether I have incorrectly defined my governing equations or if I am encountering some computational limits.
Could you provide guidance on this issue? I have attached the governing equation and my setup for reference.
Best regards, AA
Please login with a confirmed email address before reporting spam
FYI, I ran your model for some of your overall time span. It seemed to execute without complaints. I don't know what this model corresponds to physically, so I have no idea whether it is computing correctly or not. But it didn't generate an error, at least for as long as I executed it. I've attached a plot of your probe variable vs time. Note that non-linear time-domain problems can suffer from all sorts of stability issues, so even if you have everything else set up correctly (and you may not), you may still need to carefully adjust various solver settings and time stepping parameters to obtain physically meaningful results.
-------------------Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
Attachments: