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.

pure advection?

Please login with a confirmed email address before reporting spam

Hi there,

Is there a good way to implement an advection equation in COMSOL? something of the form dt(A)+dx(Av)=0 where dt and dx are time and space derivatives.

I've tried using one of the chemical species transport models and setting the diffusivity extremely low (I get an error if I set it explicitly to zero) to solve dt(A)+v*dx(A)=0 and I've entered my own PDE to solve the first equation above. For an initial square wave I get fluctuations near the wave edges that grow with time, just like I would get if I tried to write a basic finite difference approximation of a pure advection equation. Even cutting COMSOL some slack and using a wave with less abrupt changes gives rise to instabilities. Is there a solver setting somewhere that could help? Are there any COMSOL tricks to adding pure advection routines with effectively no diffusion?

Thanks!

Chris

3 Replies Last Post Jan 22, 2011, 12:32 p.m. EST
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 22, 2011, 2:31 a.m. EST
Hi

for leartning how to use COMSOL for PDE's, especially in chemistry, I can only advise to take a look at the book of Zimmermann (see the related "books" on the main COMSOL site)

--
Good luck
Ivar
Hi for leartning how to use COMSOL for PDE's, especially in chemistry, I can only advise to take a look at the book of Zimmermann (see the related "books" on the main COMSOL site) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 22, 2011, 11:34 a.m. EST
Thanks Ivar,

I've looked at the Zimmerman book before and I don't remember it dealing with pure advection. I'm also comfortable using the PDE mode and have set up several of my own equations in past projects. So I don't think it's a matter of how to put the equation into COMSOL's PDE mode.

It's a question of if I put in a PDE with only a time dependent term and a conservative flux term, how does COMSOL solve it? From my tests it seems that it tries to solve it in a similar way to an advection-diffusion problem; it looks just like it would if I tried to solve the conditionally unstable advection equation via finite differences. So what I'm wondering is if this is something I can fix. Yes, I can change the solver, preconditioning, etc., but I think this is a problem with the method in which COMSOL takes my equation and processes it to get it into a solvable form.

Thanks Ivar, I've looked at the Zimmerman book before and I don't remember it dealing with pure advection. I'm also comfortable using the PDE mode and have set up several of my own equations in past projects. So I don't think it's a matter of how to put the equation into COMSOL's PDE mode. It's a question of if I put in a PDE with only a time dependent term and a conservative flux term, how does COMSOL solve it? From my tests it seems that it tries to solve it in a similar way to an advection-diffusion problem; it looks just like it would if I tried to solve the conditionally unstable advection equation via finite differences. So what I'm wondering is if this is something I can fix. Yes, I can change the solver, preconditioning, etc., but I think this is a problem with the method in which COMSOL takes my equation and processes it to get it into a solvable form.

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 22, 2011, 12:32 p.m. EST
Hi

in the new v4 you can really teak nicely the solver sequence, but this is really starting to work correctly from 4.1 I must say. But you need often, in such cases to do some manual tweaking (things tat were far more complicated or required matlab in 3.5)

--
Good luck
Ivar
Hi in the new v4 you can really teak nicely the solver sequence, but this is really starting to work correctly from 4.1 I must say. But you need often, in such cases to do some manual tweaking (things tat were far more complicated or required matlab in 3.5) -- Good luck Ivar

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.