Re: Boundary conditions

ganesh

Dear Friends,

I am solving a 2D linear scalar advection problem with a known exact solution. The problem is similar to the one posted in Wiki, where the solutions are concentric (semi,in my case)circles. I have performed the experiment by taking a 2*2 domain, where the upper,left and right boundaries have zero values for the scalar, and the value on the lower boundary(through which the scalar enters and leaves) is prescribed from the exact solution. This happens to be a half domain( ie. the solution are concentric semicircles), and I expect to get my solution and error symmetric about the y-axis. Unfortunately, the soution is asymmetric and hence the error. I am puzzled because there is only a single boundary where I need to specify the BC, in some fashion since the values are 0 at other boundaries, but I still do not get symmetry.

Any suggestions would be helpful

Regards and Thanks in advance,

Ganesh