Dear STAR CD users,

How do we calculate the "y+ value" to assign boundary thickness?

help me out in this. ----BV

I will tell you the way that I do it. I use:

y=14*(y+)*(viscosity/(density*velocity))

Then I assume that y+ is about 75 and solve for y, where y will be my first cell layer thickness. This is actually the formula for a flat plate, but gives a pretty good estimate for other geometries also. The thing to remember is that you need to keep units the same. So if you use SI for the fluid properties, then the y value will be in meters, etc.

Tom

Thanks Tom

A more generic approach to calculation of Y+ can be outlined as follows:

Step-1: Workout 'k' nased on estimate of local velocity magnitude 'u' and tubulent intensity 'I' (k=1.5*(u*I)**2, I ~ 0.01

Step-2: Esimate turbulent dissipation rate 'epsilon' based on 'k' and turbulent mixing length 'L' (usually of the order of 0.01*Hydraulic Diameter) where epsilon= (0.09)**3/4 X k**1.5/L (0.09 is an empirical constant)

Step-3: Work out tubulent coeff of viscosity 'mu_ee' which is equal to 0.09*density*k**2/epsilon

Step-4: Estimate wall shear stress (tau_wall) based on assumed velocity profile (appropriate wall law) i.e. tau_wall = (mu_tee + mu_lam)* du/dy

Step-5: Work-out u_star

Step-6: Finally, estimate Y from definition of Y+ assuming Y+ ~ 3-5 (in case viscous sublayer to be resolved and Y+ 30~50 in case only turbulent zone is to be resolved.

(Note that this estimation of Y would be based on CELL-CENTROID location and not cell thickness).

Hope this will help! Amod Kumar

PS: Some FLUENT tutorials for example such as developing turbulent flow in pipe do suggest empirical formula to calculate Y+. However, I have not come across similar expression in STAR-CD. Has anyone come across ADAPCO recommendation to calculate Y+?

I believe the FAQ section on adapco-online has basically the same explanation.

Yes Pauli, it has been re-phrased (alongwith some recommendation about I & L)from the info available at

http://www.adapco-online.com/adapco_...vices/faq.html

I forgot to give this link.

Amod

respected sir

i want to get some information about y+.

why is this y+ important?

how do u define y+ in mixing length CFD solver?

why k-e turbulence model is better than mixing length model.

explain me clearly about both these models, and y+

Not an easy series of questions to answer. Simply, y+ determines which wall boundary conditions exist. Some texts state the near wall boundary is laminar if y+<= 11.63. The analysts have two ways to go - design the mesh to give the y+ to suit your model, or if you can not do that, choose your turbulance model and wall treatment to suit your y+. y+ is defined by the thickness of your subsurface. To determine what y+ is after the run, you must specify that the code print y+ values. Then do a getw,yplus then do a wall plot (wplot). The mixing length for y+ is usually set as one order of magnitude smaller than the hydraulic diameter. I would suggest the text Turbulence Modeling for CFD by David Wilcox. Tom