Hello, Is it possible to impose the symmetry BC for an axis-symmetric problem in cartisian coordinates? than you.

Yes, it's possible. What is the geometry and what are the flow conditions in your problem?

Hi Mani; I use a cylindrical tube with inflow conditions at the inlet, outlet conditions at the outlet and symmetry for the half domain. thanks.

Just making sure: what exactly do you mean by "half domain"? Is that really half of the 3-D pipe, i.e. you have a 3-D grid for half of the circumference? More likely you mean, that you discretize half of the radial cross-section with a 2-D Cartesian grid. Sorry to be so specific, but if you really want it to be axisymmetric, you would rather do the latter. The symmetry condition should work. Do you have any problems with that, or why are you asking? Maybe you can describe your problem in more detail.

I discretize half of the radial cross-section with a 2-D Cartesian grid. the problem that I have encountred that I have lost the symmetry in time when I simulate the motion of a bubble. perhapse, I haven't implememted the boundary condition of symmetry correctly or the grid chosen is not accurate.

The way you calculate the flow, i.e. the choice of your domain and the boundary conditions, inherently implies axisymmetry. I don't see how you can "loose" symmetry in time. Regardless of the accuracy of the solution, it must be axisymmetric by definition. If you have a solution on an x-r cut (where x is the direction of the pipe axis, r is the radial direction) for r from zero (center line of pipe) to R (radius of pipe wall), then the 3-D solution is produced by rotating your solution about the pipe axis. It is therefore axisymmetric, and I don't understand how you can judge that it's not? Are the symmetric boundary conditions not satisfied?