y+ in a train surface
 

Hi!!

I'm working with a train model and i wanted to prepare the grid in order to have a y+ close to the surface equal to 100 (because of other studies performed and capacity...) anyway, i know that the computation of the size cell should be like that:

d/2=y=(y+·viscosity)/u_tao

where u_tao is calculated as sqrt(Cf·u_infinity/2)

my question is what is Cf and how to get this?? i know there are expressions depending of the flow (laminar or turbulent) but for a U_infinity=40 m/s which one should i use???

thanks in advance!!!

Lottar

Re: y+ in a train surface
 

Hi Lottar

u+ and y+ non-dimensional forms are usually used for turbulent flows, with the definitions:

u+ = u_mean/u_star

y+ = y*u_star*rho/mue

with wall friction velocity u_star = sqrt(tau/rho)

and tau = mue*du/dy @ the wall (y=0)

The link to the skin friction is: Cf = 2*tau/(rho*U^2)

You should calculate also the Reynolds number (Re=rho*L*U/mue) to see what flow regime you have (laminar or turbulent). But as I mentioned at the beginning u+ and y+ values are mostly used for turbulent flows, so far as I know.

Hope this helps!

M.

Re: y+ in a train surface
 

Thanks Magellan!!

But in this case is difficult to know which is du/dy at y=0 and so on, so the best way to get y+ i found was with

Cf=0.0588/Re^(1/5) since a turbulent boundary layer as u say will appear and the surface on the train could be considered as a flat plate...

what do u think guys?

cheers!

Re: y+ in a train surface
 

Hi Lottar,

Just make a coarse approximation, create a grid accordingly, perform a simulation and adjust your grid based on the computed and desired y+ values.

Re: y+ in a train surface
 

Ok, i'll try in this way...thanks Timon!