hi,everyone!i meet one problem:i can't use periodic boundary conditions in multispieces simulation in fluent when using coupled solver.i have tested all the conditions used in my models.last i simpfied my problem as follows for us to discuss:

the region is a big rectangle(the diagonal points are (0,0) and (10,6)) subtract a small rectangle( the diagonal points are (2,1) and (8,5) ) ;the edge line0,0)-(0,6) is used pressure inlet boundary condition;the edge line(10,0)-(10,6) is used pressure outlet boudary condition'the edge lines0,0)-(10,0)and (8,0)-(10,8) are set as periodic conditions(translational,no pressure jump);the fluid materials are n2 and o2(any kind of mixture,and material property can be set as any kind of choice) ,when i use coupled solver,fluent reports errors(just be helpless)!

i read the users mannul again, and don't find any questions in chapter6(chapter8 is not my case).

any suggests are welcome!

thanks

ydzhang

Are the meshes on the periodic boundaries linked?

thanks Kuochen for response!

of course i link the meshes.i have test making periodic boundary condition using gambit and fluent(mp command),but it still can't work! ydzhang

My experience shows that the coupled solver won't handle periodic boundary conditions at all. I did some 3-D simulations with finite-rate kinetics with the coupled solver. By symmetry, only a quadrant is needed, but I have to use the full geometry in order to obtain convergence. My advice is to avoid periodic boundary conditions by all means. It does work for symmetry boundary condition though.

thanks,K. Tsai yes,you are right!now i am also using symmetry or wall to avoid the problem,and it does work! but my situation is that i must use periodic boundary conditions to discribe my problem.i want to write UDF to solve it,i am don't know how to realize the periodic boundary condition in my UDF.any advice or message is appreciated! ydzhang