BC's don't work well

Log Out | Topics | Search Moderators | Register | Edit Profile

Author Message   Olexiy Orel (scorpion) Member Username: scorpion Post Number: 15 Registered: 03-2007 Posted on Friday, February 05, 2010 - 07:18 am:    Dear sirs, I have the following problem. I set BC's on anodes: i(c)=-C*alpha, where C is the concentration, i=natural(C) is a current. Anodes are of spherical shape. When I look at elevation(C) on anodes and elevation (normal(grad(C))) on anodes - they are not alike. BC do not work well all the time! Please see attached files. One needs to run sperical.pde file, then Navier-Stokes.pde file. Then .dat- files created by them are used in C1.pde file. first files, makes stationary concentration spherical.pde (2.3 k) second file, makes velocity file Navier-Stokes.pde (2.0 k) main file Ñ1.pde (2.6 k)   Olexiy Orel (scorpion) Member Username: scorpion Post Number: 16 Registered: 03-2007 Posted on Friday, February 05, 2010 - 07:21 am:    so, once again. I have anode of a spherical form. And I put BC on it: natural(C)=-alpha*C. Then I see elevation of C and elevation of normal(grad(C)) on anode and they do not fit each other. Where is the trick? May be natural(C)==normal(grad(C))? Thank you dear sirs, you are so kind.   Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 1325 Registered: 06-2003 Posted on Friday, February 05, 2010 - 02:19 pm:    You have a circular file dependency here: Spherical.pde depends on Navier-Stokes output. Navier-Stokes depends on Spherical output. Please correct this and re-post. Is there a reason why you do not solve all these together in a single script?   Olexiy Orel (scorpion) Member Username: scorpion Post Number: 17 Registered: 03-2007 Posted on Friday, February 05, 2010 - 03:15 pm:    Dear Robert, I've fixed a circular dependency. I upload the first file again. The only reason is that I want to save computation time. It is not necessary to calculate stationary concentration each moment. I calculate it once and then use the result in another file. spherical.pde (2.2 k)   Marek Nelson (mgnelson) Moderator Username: mgnelson Post Number: 180 Registered: 07-2007 Posted on Friday, February 05, 2010 - 06:42 pm:    I think the plots at the anode boundary are in the range of numerical noise. Values of C1 at the boundary are on the order of 1e-6 but the range of value within the system is about 1e-2. Also, the resolution at the anode is also limited by the grid size. If you increase the grid density by putting mesh_spacing=R/100 on the anode/kathode, the plots will show the same shape even though they are noise.