Dear Sirs,
Is there an easy way of extracting averages in an homogeneous direction (z, for instance) for some resolved variables?
Tnaks,
Luiz

[maka]

I'm trying to make the channelOodles average channel data in the x and z directions (periodic). I tried to modify Probe.C to make SpaceAverage.C.

I came to the point where I need to know how can I call:
(1) (x, y, z) of all cell nodes
(2) loop over all node values of U, p in (x or z direction).

I would be grateful if you can point to a part of the code where I can learn how to access the data structure to achieve such objective.

best regards,
Maka

[eugene]

Hmm, you might want to try "postChannel". Not sure exactly what it does, but I think you might find it useful.

[maka]

Thanks. This is just what I needed.

Regards,
Maka

[anne]

Hello,

I have used posChannel and I am very happy with.
HOWEVER, it is written for two homogeneous directions and I want now to make average for a
backward facing step flow which has only one
homogeneous direction.

So I would like to know if there is any function average
that operates the average in only one homogeneous direction.


I saw that there is a average fuction used in
the LESmodels sources but I am not at all sure bout what this function does: does it recognize
from the polyMesh directory the cyclic BOundaries and
thus applies only the average in the homogeneous directions ?


I have tried to extrapolate the collapse function used in postChannel but I am not sure that
the compuation of the indexes is general and can be applied to my case.

I saw on the forum that people use some visualization tools to do so but I am interested
in computing it myself to then make comparisons with data or renormalize the other data on my own.


I would be very please if you could orientate me in this task,

Thanks

Anne

[cedric_duprat]

Hi forum,

I would like to re-open this tread to answer my previous question from here: http://www.cfd-online.com/OpenFOAM_D...es/1/2946.html

now, postChannel is very efficient to post-processes data from channel flow calculations. so, it assumes that the mesh is periodic in the x and z directions.
I want to create a new postChannel for just one periodic directions (let's say z).
This can be usefull for more complex geometries, to get rms values at different distances from the inlet. I think we could also use it for periodic pipe (I think ...) or axisymetric geometry

So, in postChannelDict, we should be able to add a list of x value(s) and not a number of cell. This list can be one or more value(s).
The member Functions "collapse" will do a "collapsing" of all the value in the Z direction to get a single y-line for each plane x of the list.

My questions are how to define the list in postChannelDict, how to read it in the "indexingDict_" (channelIndex.C) and then, how to loop over this list. Because the list is not supposed to be a constant x-step list.

I think we can read only one value of x which could be defined in channelPostDict but, if we want to do more than one rms, a list should be more efficient.

This can be generalised for zero periodic mesh, maybe a "value" in postChannelDict can switch the solver from one method to an other. The idea is to do something less focus on channel flow and more usefulll for complex geometry.

Can someone on the forum tell me what he is thinking about a tool like that ?
Maybe someone already did a tool like that before ? why not :o)
Could it be usefull for someone else ?

Well, ... that's all.
Thank you for all in advance.

Cedric

[agustinvo]

Hello

I know this is a very old thread, but I would like to know if you were able to do this. I will use something similar for my application (a boundary layer).

Thanks in advance!
Agustin

[nukecrafts]

Originally Posted by agustinvo

Hello

I know this is a very old thread, but I would like to know if you were able to do this. I will use something similar for my application (a boundary layer).

Thanks in advance!
Agustin

re-surfacing issue, if anyone has done it