Meshing strategy for a tapered tube

Dave

Hi all, I am relatively new CFD. I would like to model a laminar pulsatile flow in a 2d axisymmetric conical tube using Fluent 6.1.22. This tube can be seen as 3 seperate parts; Inlet is constant diameter pipe d1, outlet is constant diameter d2 width d2>d1. The intermediate section is tapered.

I have attempted to mesh this tube in 3 separate sections (described above) and used a boundary layer on the only solid wall boundary to refine locally the grid.

After having looked at severall papers it appears that grid refinement study is very important. Refining the grid in 3 seperate sections seems to require a "trial and error process". This problems seem worst due to the presence of the boundary layer at the wall.

Is there a better strategy for meshing such a conical tube in light of the need to refine the grid for grid sensitivity studies?

Best Regards, Dave