Discussion Forum

Dear Nils,
I think you'll find what you're looking for is a "tickable" box that appears in the Settings window for the "Transport of diluted species" module. Under "Transport mechanisms" you may select convection and migration. I hope that helps.

By the way, I am interested in a similar problem and I am having trouble solving the NPP equations under transient conditions: have you succeeded in that? what study configuration have you used? I have tried several things but none seem to work.

Cheers,

Javier.

Dear Javier,

thanks, I found the solution even before you said, but thanks anyway.
I have only done diffusion in transient problems, and I learned (trial and error) that you need to have a certain ratio of your mesh size and time step. In general your mesh cannot be too large for a given time step (otherwise diffusion cannot occur between adjacent elements).
Also, you might wanna check which times you are simulating. Is it all times that give weird results or just for small ones? I had a problem of not getting right results for small times and discovered that the program would do large time steps (as defined by the largest time I simulated) for all times. You can circumvent the problem by defining time steps your own (in the solver settings) and by setting an additional time 2 orders of magnitude lower than the lowest time you're interested. Otherwise comsol will start with the lowest time as initial value and you get only crap.

Hope that helps.

I took a step backwards, and am trying to model an electrical double layer, next to a charged surface, however I only get error messages (maximum no of iterations reached, returned solution has not converged). Did you do anything like that? Do you have any tips? I combined electrostatics and transport of diluted species and set all the boundary conditions.

Thanks
Nils

Hi

For the time stepper tweaking

Do you not have the "Solver - Time Dependent Solver" Node - Time Stepping - Steps taken by the solver field ?

In which case you can select strict (= what you define), intermediate (= Comsol adds an extra point in between yours), and automatic (by default = Comsol governs alone, and at best, but not always optimum)

--
Good luck
Ivar

Hi Comsol community:
I want to simulate the ionic current (KCl as buffer) value through a 2-D rectangular nanochannel under voltage bias. I only consider the effect of diffusion and migration (I dont care about Navie-Stokes since I am not solving for fluid velocity field or convection), and more importantly how different surface charge on the wall have an effect on overall ionic current through nanochannel.
This sounds simple but I could not find a good model in the gallery or simple tutorial online.. Could someone give me a hint as a head start in COMSOL's 4.1 version? Thank you very much.

Use the diluted species module and electrostatics in the AC/DC module.

Hope this helps

I am trying to solve a similar problem - the formation of an electrical double layer over time, using Poisson's equation and transport of diluted species.

I am trying to input the concentrations calculated by the TDS interface as the source term in the Poisson equation, but as the Poisson equation is time-independent it is only calculating the electric potential at time zero, using initial conditions. Do you know if you can force Comsol to solve Poissons equation at each timestep, with the updated concentrations? Or is there any other way of calculating how the electric potential changes as the ions migrate?

Thanks,

R

I hope you to upload your model.If you just only asked something like Nils but offerring no more,it is no helpful to all others.

Let's share each other and progress together!

Dear All,
I am new user to COMSOL. I am using COMSOL 4.3b version and trying to couple Nernst Plank equation and Poisson's equation for conical quartz nanopore employing 2-D axisymmetric space dimension. Do you have any idea, for this model, whether I need to draw full geomotry of nanopore or only half section? I drew full nanopore geometry using "Bézier Polygon 1" and finally converting all these to solid.
When I changed the bias polarity, I didn't find any change in solution. What may be the possible cause for this?
Thank You

Hi purushottam,
You dont need to draw a full geometry of the your system in 2D-axisymmetric. Draw your geometry with half of the dimensions of your geometry and you can see your full geometry after computing in results section.