Geometry for my case * cylinderical coordinate R: 0 ~ 10cm (Nr=100) THETA: -1deg ~ 1deg (Nt=1) Z: 0 ~ 10cm (Nz=100)

Therefore, I used the symmetric boundary for THETA=-1DEG plane and THEAT=+1DEG plane. But the result is stange at axis symmetric line(R=0). In this case, Do I use the cyclic boundary for THETA planes?

Did you do a VMERG? The collapsed cells along the axis must have the correct structure.

Also try CFLIP PRISM to ensure the prism cell numbering is correct.

Cyclic boundaries should be used when there is flow across the boundary, i.e. when you have tangential flow. If there is nothing to induce flow in this direction, use summetry boundaries.

I did VMERG.

I used cyclic boundary with considering 3-D and intial value U,V,W=0. and.. I got the resonable results. Thank you, Brian.