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 The information provided may be out of date.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.
Flux units in surface reaction interface module, and modeling of surface diffusion in axi-symmetric system
Posted 2013/03/07 13:31 GMT-5 Chemical Reaction Engineering Version 4.3a 3 Replies
Please login with a confirmed email address before reporting spam
I am trying to model a bulk - surface reaction-diffusion system (diffusion in the solution, diffusion on the surface, and adsorption/desorption which couples the two together).
The system has a cylindrical symmetry so I want to solve it in an axi-symmetric configuration, and I have two questions regarding the implementation of such system:
1. Initially I constructed the system similarly to the example of Transport and Adsorption case study (Model Library > COMSOL Multiphysics > Chemical Engineering>Transport and Adsorption) which is quite similar. In this case the surface is modeled using a General Form Boundary PDE (gb). I found out that the surface concentration was not solved correctly, and I suspect it is related to the axial symmetry.
Did anyone encountered such a problem?
Does general form PDE work with 1D axisymmetric configuration?
2. I rebuilt the model with surface reaction interface instead of general PDE, and this indeed gives a correct solution for the surface concentration.
I want to check a flux on a point along the surface using the total flux magnitude in the post processing (chsr.tfluxMag_). However, it reports it with the weird units [ mol^2/(m^3*s) ] ,which are definitely not flux units.
The value is also not in agreement to the predicted flux (in fact it is the correct flux*constant).
What is the "flux" which is given there?
Is there another way to calculate the exact flux in this interface?
Thanks!
Henry
Please login with a confirmed email address before reporting spam
could it be that you are confused about the 2D or 1D normalisation: COMSOl always calculates in 3D, the 2D is a by default 1[m] depth 3D (constant along Z) so you your units are all normalised for 1/m, while 1D is normalised per Area hence a 1/m^2. You resolve this by inteegrating over the surface ( line in 1D) which represents the "domains" and then multiply by the depth "d" or Area A and your absolute units come back.
In 2D axi its 2*pi*r, the loop length that is the normalisation value, in 1D-axi its 2*pi*r*height ...
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
This can solve some of the issues and I did take it into account, however it still doesn't solve the problem.
When integrating over the loop (this is an axi-2D problem in this case) you still have units of mol^2/(m^2*sec), and even if I assume it is a 1D problem I get mol^2/(m*sec) instead of mol/(m*sec).
I am mostly puzzled by the mol^2, because this can not be related to the normalization issue.
Where can this come from? I rechecked and the surface concentration is defined correctly (as mol/m^2) so it is not the problem.
Thanks,
Henry
Please login with a confirmed email address before reporting spam
Sorry I missed the mol^2 and jumped on the denominator as that is where one usually forgets the nrmalisation.
Indeed its uzzeling, would have to dig into the formulas again, to try to understand
If no one answers, try support ;)
And keep us posted
--
Good luck
Ivar