u_tau and tau_wall don't have to be arrays

[Niceno]

Dear guys,

I was working in implementation of multi-domain (and later multi-material) capability and have noticed that two arrays; u_tau and tau_wall don't have to be arrays at all. They are used in four functions (Src_Kin_K_Eps, Src_Eps_K_Eps, Src_Kin_K_Eps_Zeta_F and Src_Eps_K_Eps_Zeta_F) but are always re-computed, never passed from one to another.

I suggest we turn them into normal variables defined locally in each of these functions. Is this OK for everyone?

[mohamedalysayed]

I agree with that it shouldn’t be a “global” array but still be a local one browsing through the boundary cells (since tau_wall can differ from one boundary cell to another in irregular geometries). What do you all think ?

[Niceno]

I think local variable would be sufficient since it doesn't store values in boundary cells, rather in cell inside the domain.