Discussion Forum

If a single solute transfers from the inner sphere through a wall to the exterior of the sphere, you do not need to specify any BC at the inner surface, just define the diffusion coefficient in the inner sphere and in the sphere wall. You need thus only one Transport of Diluted Species node.

BR
Lasse

Thank you for the suggestion sir,
But, See the image attached.
I have tried what you have suggested.
Here I A have 2 concentric spheres (say s1 and s2 starting from the inner most one) over which a fluid is flowing . I have applied your suggestion of 2 diffusion coefficients in the 2 domains and on the outermost surface I have used a flux condition as fluid is flowing over.
my doubt is what condition should I specify at the inner most boundary of s1? 'No flux' condition is not working in my case. There my concentration is a function of time. I have searched a lot but I didn't get any appropriate analytical expression for time dependent concentration.
Can you provide any help in this case?

I thought that the inner sphere is the source of the diffusing species, but in your image that is an empty cavity, i.e. it is not a part of the simulation domain?

If you need a time dependent concentration at the inner boundary, define an appropriate function under

Component --> Definitions --> Functions

Let's assume that you have defined a piecewise function that is named "pw1". Then, at the boundary call it as pw1(t).

BR
Lasse

Thanks for the suggestion sir.
Sir,
In my model the thickness of the 2 layers decreases with time because of diffusion. So I have used the deformed geometry physics for this purpose.
In my inner layer the concentration is constant and becomes zero only when the thickness of this layer becomes zero.
When I am running the simulation with deformed geometry. It was displaying an error, 'Inverted mesh element'
Do you have any suggestion about how can I over come this?

When the thickness approaches zero, there won't be anything to be meshed. I have no clear solution to this, but if you define the thickness by a parameter, also the mesh can be parametrised. Like:

d = thickness of the layer
N = d*10 = number of layers in a mapped mesh

What kind of meshing are you using?

Thanks for the suggestion sir.
I am using free triangular mesh.

Sir,
Suppose my concentric spheres are oppositely charged, is there a way I can incorporate this in my comsol model.

Hi

This is a question I have been thinkng a lot because I would like to model ion-exchange membranes that have fixed charged groups. So far I have not figured it out how to do it, and I have not found anything from this website or in the web.

BR
Lasse