[mostanad]

Hello everybody
I have solved a duct with my fortran code, so I wanna validate the results. thus my main questions are:
exactly what is the correct way for post processing following terms?
1- I need shear stress of the surface so I calculate velocity gradient with upper cell. my problem is which upper cells. the cells that adopt their numbers with surface cell numbers? or closest one?
1- I wanna validate law of wall. so I need u_tau (friction velocity). I have friction velocities of surfaces. but what is the friction velocities of upper cell centers in domain. I know it is the amounts of cells on wall but for node in j=4 and i=30(4 node upper the wall) is friction velocity for j=1 and i=30 correct? or I should use the amounts for closest cell?
I will be so happy to help me or introducing a tutorial.

[mostanad]

No answer???? really!!!

[FMDenaro]

If you are solving DNS or LES with resolved BL, you can also adopt a simple first order accurate forward/backward normal derivative. Second order accuracy requires three nodes (one on the wall, two in the fluid region).

The velocity u_tau is that defined by tau_wall = rho*u_tau^2

[mostanad]

thank you Mr Denaro,
my main question is about u_tau in domain, that is defined by the wall. imagine in two or three nodes over wall, what is the amount of u_tau???

[FMDenaro]

Originally Posted by mostanad

thank you Mr Denaro,
my main question is about u_tau in domain, that is defined by the wall. imagine in two or three nodes over wall, what is the amount of u_tau???

that is not clear to me

[flotus1]

You calculate u_tau for the wall once. For the analysis you (probably) have in mind, this value of u_tau has to be the same for all distances from the wall.

[mostanad]

Originally Posted by FMDenaro

that is not clear to me

in the domain we have a wall and many grids above it. i can find the u_tau for the wall. this amount can be changed along the duct. so what is the amount of u_tau for nodes above the wall. because i wanna define y_plus and u_plus for all nodes in domain.

[mostanad]

Originally Posted by flotus1

You calculate u_tau for the wall once. For the analysis you (probably) have in mind, this value of u_tau has to be the same for all distances from the wall.

exactly flotus1,
but if we have various u_tau on wall and also displacement in nodes in normal direction so how you relate u_tau amounts on wall to upper grids? as you see in pictures we have four nodes. so we have u_tau for them. u_tau changes from 1 to 4. but for nodes in far from wall if we have displacement, what is the u_tau amount?

[flotus1]

Use the value from the nearest wall node. For higher accuracy you can interpolate the values of u_tau to the wall position with the closest wall-normal distance to the cell itself.
utau_wall.jpg

[FMDenaro]

There is a unique velocity scale in the inner region

https://www.cfd-online.com/Wiki/Intr...urbulent_flows

[mostanad]

Originally Posted by flotus1

Use the value from the nearest wall node. For higher accuracy you can interpolate the values of u_tau to the wall position with the closest wall-normal distance to the cell itself.
Attachment 52934

It means that i should calculate the distance from walls(exactly from it's nodes) and find the nearest wall distance, also that node in wall?

[aerosjc]

\U_tau is defined at the wall. If you've already calculated it , you can further calculate y^+ for any node. y^+ = y \nu / u_tau.