[mazdak]

Hi dear readers.
as you know when you use finite volume method, for turbulent flow,you need to fix k and epsilon at nodes that wall function are applied trough below method:

Su=(10^30)*...
Sp=-(10^30)

I wanna to know when we use a collocated grid do we need this treatment for U-velocity when we use wall-function?

and is there anyone write a code for turbulent flow in channel in a collocated grid?

mazdak_parsi2000@yahoo.com
I am waiting for your helps.

[hadian]

you should calculate the U at the nearest cell to wall and add the shear stress of the wall to the source term for that cell. there are some tricks for adding the stress. have a look on the book by Versteeg and Malalaskera.

[mazdak]

thanks Mr.hadian but what you say is for staggered grid nor to collocated grid.i study that book.thanks

[hadian]

do you mean the u-cell near the south or north wall? so it is the same as staggered grid. for v-cell near north and south wall the method is a little different.

[mazdak]

imagine an horizontal channel.as you know, when you use collocated grid, all variable u,v,p,k,ep calculated at same cell.

as you study versteeg's book, you know that because some nodes are outer of walls.i mean we have not any nodes on wall for staggered grid.so the effect of wall, modeled by adding a source term to Sp in u-equation.

in this added source term u-plus is available.if y-plus be less than 11.62 we treat it, in a way and if y-plus be greater than 11.62 we treat it in another way.

my question is when we use collocate grid how do we treat for wall function?

if you are live in iran and you know any thing about using wall function for collocated grid i call you

[hadian]

in collocated grid you have a cell out of the wall. that is exactly the same as staggered grid.
www.hadian.ir

[mazdak]

Hi Dr.hadian

i didn't know really this point.but the generated by me has this properties:

imagine a south wall. i put first row of nodes on wall.second row of nodes puted upper than those and after second row, i put north cell face of cv.
like this

------------(north cell face of cv)
................. (second row of nodes)
................(south wall nodes)(first row of nodes )(south cell face of cv)

i mean i don't put any nodes out of wall and use this form of grid to form complete cv.
what's your opinion about this type?

[hadian]

because of boundary condition you should have extra nodes in your domain. one method is to have dummy cells out of boundary (as it is described by Vesteeg and Malalaskera) the other one is to pot extra nodes on boundary. in the second form you do not have cells beyond the domain but there will be some cells with more than one node (one at the center of cell and one on the cell face). i think the first method is more straitforward.

[mazdak]

Hi dear Dr.Hadian

i choose 2nd form of grid that you mentioned.question is about imposing boundary condition for u velocity and utlizing of wall function.i use
Rhi-Chow for eliminating pressure hardship when we calculate all variables at same node.and if you can send your cell phone number in my email address to call you.

[mazdak]

Hi dear Dr.Hadian

could you tell me how must i treat for the V near the wall in collocated grid?

and if you have any paper in this field pleas send me.

be in success.

[hadian]

you can calculate the V at cell near south or north wall in your calculation. for calculating the coefficients you will refer to values at the cell beyond the wall. after calculating V at cells. use -V of the cell next to wall for the cell beyond the wall. In this way V=0 in implemented on the wall automatically.

[jyothishkumar]

Dear Hadian,

I have cfd code for laminar flow (channel flow) with colocated grid arrangement. I have implemented k epsilon turbulence model in to this. Now if i see my velocity profile (coming as expected) i found that it is more peakier than that of a laminar profile. Also the pressure at the inlet increasing like anything. Details are as follows

My boundary conditions : Velocity inlet, pressure outlet and wall (top and bottom).

And I have followed the book versteeg and malalasekara. Also I have compared my laminar flow results with FLUENT code and found it is matching but i am getting this peculiar problem for turbulent flow. Please tell me your valuable suggestion.

thanks in advance

jyothish