domain size for Detached Eddy Simulation

T

Dear all, I am modelling a series of horizontally placed circular rods (bars) placed in uniform incendent flow conditions with DES. In my model 10 rods are included and I use symmetric boundary conditions in the spanwise direction and periodic boundary conditions on the top and the bottom faces of the domain in order to simulate an infinite series of rods instead of 10. I wondered if I have to make the domain (the number of rods I actually model - in this case 10) large enough so that it is at least as large as the largest eddy in the incendent flow with DES which is dependent on the length scale of turbulence in the incendent flow. If this is not necessary I could only model one rod with periodic boundaries at top and bottom but can DES solve the larger eddies correctly in this situation by using periodic BC's? Thx