Impressions of V 6.3

Allan Walsh

Now that version 6.3 has been released for awhile, what are people's impressions?

We have played with it a bit and so far have not found any compelling reasons to move from V6.2. We are modeling boiler combustion systems with a generic set-up - segregated, implicit, steady-state, species transport, k-epsilon, etc.

The hexahedral grid is of limited use to us. We use fairly coarse tetrahedral grids and then use the grid adapt feature in Fluent to improve the grid resolution in the regions of high velocity. Fluent 6.3 does not allow converting adapted tets to hexes. Or vice versa, hexes cannot be adapted - so this hasn't been too usefuel.

Under Windows, the graphics had been broken going from 6.1 to 6.2. The pathline function would crash and the sweep surface was verrrrrrrrrry slow. These seem to be fixed in 6.3.

Reading cases with UDFs into different versions of Fluent is ugly since the UDFs all have to be recompiled. Then, the variable name for particle position was changed between 6.1 and 6.2 (or was it 6.0 and 6.1?). And, allowing Fluent to reference multiple dlls has caused problems with linking UDFs.

We used version 3 of Fluent quite extensively in the 80's, V4 in the 90's, not much of V5, and now use V6 on a daily basis. For our applications, V6.3 does not seem to offer much new.