Menu

Discussion Forum

New Discussion
Filtern

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 The information provided may be out of date.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.

Concentrtion gradient in Time dependent simulation at time t=0

Posted 07.10.2014, 02:04 GMT-4 Chemical Reaction Engineering, Studies & Solvers Version 4.3, Version 4.3a, Version 4.3b 3 Replies

Kishore Gajula
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

I have a system in which diffusion takes place from a sphere to a medium(in a cylindrical domain)..i am running time dependent simulation using "Transport of diluted species " physics.
Initial concentration in sphere = C0(some constant value) at t=0
outside sphere = 0 at t=0
1) After the simulation when i checked for concentration profiles in the sphere at t=0 ...it is showing some concentration gradient at the end of the radius of the sphere.Ideally there shouldn't be any gradient at t=0 inside the sphere since i gave some initial concentration(t=0) in sphere and zero concentration outside it.

can anyone tell me how to resolve the issue??
i tried specifying some flux at interface ...but it wouldn't work

Thanks
Kishore Gajula
3 Replies Last Post 07.10.2014, 05:55 GMT-4
Lasse Murtomäki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 07.10.2014, 02:49 GMT-4
At t = 0 the concentration gradient is - in principle - infinite, which numerics cannot obviously handle. But I would not scratch my head too much for that. What I usually do in a problem like yours is that I do the simulation with an exponentially increasing time step. For example, in the simulation time domain I put 10^{range(-3,0.1,3)}. In the very beginning I calculate with a very short time step, because there the concentration changes are the most rapid. I do not have point t = 0 at all, because it is a singularity.
At t = 0 the concentration gradient is - in principle - infinite, which numerics cannot obviously handle. But I would not scratch my head too much for that. What I usually do in a problem like yours is that I do the simulation with an exponentially increasing time step. For example, in the simulation time domain I put 10^{range(-3,0.1,3)}. In the very beginning I calculate with a very short time step, because there the concentration changes are the most rapid. I do not have point t = 0 at all, because it is a singularity.
Kishore Gajula
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 07.10.2014, 03:14 GMT-4
Thanks for the reply.

In my case i have to vary the time step linearly not exponentially.
I tried to simulate from t=1e-10 instead from t=0 to certain time t.still i am observing some concentration distribution across the barrier at t=1e-10 which is better than previous simulation in which there was some concentration distribution across the barrier(to end of the radius) at time t=0.

Thanks
Kishore Gajula
Thanks for the reply. In my case i have to vary the time step linearly not exponentially. I tried to simulate from t=1e-10 instead from t=0 to certain time t.still i am observing some concentration distribution across the barrier at t=1e-10 which is better than previous simulation in which there was some concentration distribution across the barrier(to end of the radius) at time t=0. Thanks Kishore Gajula
Lasse Murtomäki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 07.10.2014, 05:55 GMT-4
You can simulate in any time pattern, and present the data at interpolated values.
You can simulate in any time pattern, and present the data at interpolated values.

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.

Suggested Content