Turbulence transition region and how to model
 

I've seen some posts on similar topics before but i've yet to find a satisfactory answer (if there is one at all) to this.

How do you determine whether or not a flow is turbulent or laminar and whether or not turbulence modelling is valid?

The main problem is that the geometry being modelled is complicated enough that a valid length scale for the reynolds number is difficult to determine. Based on various attempts though it falls within the region of 1k-10k. Additionally there is a large recirculation zone that results in flow separation at the boundary and a large stagnation region where flow is negligable.

The separation point has a significant effect on the size of the stagnation region which is of particular importance so determining the best method of capturing this is important.

Any ideas?

it is never easy. you can check in the literature for previous stuides or testcases, look out for benchmarks, etc...

you can even run an experimental LES simulation, but I'd consider this as a last resort strategy because even before starting you should know what is happening inside the flow.

how big do you esteem to be the laminar region? can you share some more information about the case (what is it? boundary conditions? etc)