Boundary layer modelling can effect results dramatically, and in order to get accurate results it is necessary to get as much of the wall in the desired Y+ range. However with problems that have large differences in velocity, or transient simulation where velocity alters over time, this task of re-meshing to get good Y+ values becomes very arduous and often doesn’t bare the required results.
A solution to this would be to create an enhanced wall model which can model the boundary layer all the way from the viscous sub-layer, buffer region and log layer. Taking the example of an air foil, typically getting correct Y+ values over an entire surface is very hard and impossible towards the trailing edge unless the face is split. However, with the above method the wall could have a universally small Y value which can be treated differently depending upon its Y+ if it is anywhere within the boundary layer.
This will reduce the iterative process of meshing to get a good Y+, make good results easier for less experienced users and increase the accuracy of any validation against experimental results.
Cheers,
Darren