Efforts Directions ??

[ibrahimyaacob]

Hello

I'm still new to CFD and my first code is this Code_Saturne. This question might sound silly. I am trying to calculate the skin friction coefficient in ship hull. The formula i should use is Cf = tau/(0.5*rho*U^2).

Tau is the local wall shear stress. I believe the formula is total shear force in x component over the surface area of the ship hull. I still have doubt in the formula.

In code_saturne, i have manage to calculate 3 components of force (efforts) which are the normal effort and two tangential efforts. This efforts are applied in each every cell on the wall, while I only needed 1 value.

So my question is, can somebody please clarify to me which efforts I should use and how do I integrate the efforts in each cells to become 1 value.

Thank You

[Yvan Fournier]

Hello,

You need to go back to the definition of the skin friction coefficient.

Using the efforts and face normals, you may integrate the efforts (or their components over a given direction), then divide them by the reference surface to obtain the friction coefficient.

This type of operator could in the future be implemented as a sort of "macro" in the post_utils module, but this is not the case yet...

Regards,

Yvan

[ibrahimyaacob]

Thank You Yvan for replying

I now have a clear idea how to obtain local coefficient of skin friction. I will use the local shear stresses in every cell in the boundary. How do I obtain the local shear stress using the local efforts? Should I divide it with the area of one cell?

and I still did not understand is which one of the three efforts should I use? I realise that the efforts consist of one normal efforts and two tangential efforts, but I am still terribly confused about this.

Thanks & Regards

Ibrahim

[Yvan Fournier]

Hello,

Which of the efforts to use really depends on what you are computing (again, refer to the definition).

In any case, all surface efforts given by Code_Saturne are given per surface unit, so you don't need to divide them by the surface (but rather multiply them by the matchin face surfaces to integrate them). Also, be careful, the normal component is based on the wall-normal pressure. As the pressure in incompressible flows is not a physical pressure, but a pressure relative to a chose reference, local values will depend on your reference pressure (but on a closed surface, the integral will balance out).

Regards,

Yvan

[ibrahimyaacob]

Hello

Thank you Yvan for replying, so it is true that the normal component efforts are based on the pressure. I also got a tip from my lecturer that to find skin friction coefficient i will need to get the resultant of the two tangential component (finding their magnitude) then use them to get the local shear force.

again I'd like to ask you, from the local shear force, how do iI get the shear stress?

**********
My theory is that i must find the area of a single cell to use in the formula

Local shear = local resultant tangential effort / cell surface area

but then again I dont know how to find the single cell surface area
**********

Thank You & Regards

Ibrahim

[Yvan Fournier]

Hello,

The array providing the boundary face surface vectors (whose lengths are equal to their surface) is surfbo (in Fortran). You can find more details in the user examples (such as cs_user_extra_operations-energy_balance.f90) or Doxygen documentation .

Regards,

Yvan

[ibrahimyaacob]

Thank you I'll look into that

[ibrahimyaacob]

Hello Again

Yvan, I already take a look at the subroutine. I have no idea what or how to do it especially to input surfbo command. I also cant find any source or example to do this.. please help

Is there a way to give out the surface area using Paraview? I'm much more proficient in Paraview than using FORTRAN.

Regards

Ibrahim

[Yvan Fournier]

Hello,

We do not have many examples for efforts, except for the file I mentioned. We need to improve the documentation, but I can give you no schedule.

Using ParaView may indeed be easier for you. In this case, you need to use the "extract block" filter to select the boundary mesh (assuming you kept the defaults and have output it). The efforts are "per unit area" fields, so under ParaView, possibly combining it with surface vectors, then using the "integrate variables" filter, you should be able to compute the integrals you need.

Regards,

Yvan

[ibrahimyaacob]

Hello again

Yvan I appreciate your kind response but still havent got a clue on how to find the surface vector you are talking about. I already looked into the cs_user_extra_operations-energy_balance.f90 but there is no surfbo found.

and for the efforts, i just found out that for the flow across cylinder, the efforts are presented in dimensionless vector form and you did say that the efforts are Newton per surface unit right. so to get the shear i need to find the area of the surface unit.

For days I've been searching a way to find the surface area and I still havent got any clue yet.