Hi, While evaluating the perf
 

Hi,
While evaluating the performance of wallFunctions with roughness corrections a made some runs with simpleFoam on a really long and narrow domain; 1000m long, 100m high and 10 m wide (I want to build up an atmospheric boundary layer, I know it would have been wiser to use cyclic boundaries or perhaps a modified version of boundaryFoam, but I did it the lazy way).

The problem I get is that about 1700 "subiterations" are required for the pressure term during every main iteration. I have done similar tries with Fluent before without experiencing these problems. Does anybody know which of my setting that cause problems (I use the default settings and upwind differencing) and does anybofy have a suggestion on what to change to speed up the calculations?
Thanks
David