hi,

i am simulating turbulent flow over a ellipse and rectangle. i want to know how to calculate the Turbulence Intensiyt(I) and what length scale to adapt. i am using fluent 6.2 software. also how to compute and meshing strategy for the wall Y+

turbulent intensity is given by standard deviation/mean velocity,

where do u have to specify this number , i think it will either at the entrance or and exit ,depending on your domain , keep the value low at the exit to prevent the back flow , generally turbulent intensity is given very small, but yes this quite depending the case u r solving

can u tell me whats your geometry ?

Geting y+ is purely a trail n error method or a smart guess that comes from experience , or if u get a very simple geometry probably u must be in a position to get the estimation fo y+,

Let me know have i answered u properly or not

thnq for ur reply,

my geometry is 2D, ellipse with max .length of 26m height of 11m.

i want to specify the turbulence parameter at inlet velocity,

also for y+ should be within 30.

i am using fluent software to simulate external flow over an ellipse of major dia 2m and minor dia 1m. to calculate the drag and lift coefficient.

is the flow over the the major axis or minor axis, i mean the placement of the geometry with respect to the flow

knowledge of turbulence parameters at the entrance will come from the wind tunnel experiments , generally it is this way, but as far my experience is concerned , keepin the intensity about 0.001 should be ok , n for the outward it should be less 0.01% or something , or less enough to aviod the back flow at the pressure outlet

i hope this helps u ,,

other people reading this suggestion could u please put in suggestions over my comments

major axis

I've read the thread and I always put a turbulence int. of 1 for external flows at the inlet, and the same for the pressure outlet (in fact, this data are provided in the case of backflow, so the backflow is previous to the using of this data at the outlet).

I recommend, in any case, to take some well-known cases and try to get the same results ( I mean, simple geomteries whose drag and lift are tabled)

regards

ok, thankyou,

i will try that.