[ReeKo]

Howdy everybody,

i'm facing a truly embarassing problem when exporting solution data from fluent. I need face area magnitude of every face on an existing mesh boundary.
Having properly read the values in matlab, their sum do not coincide with the whole area value of the same mesh boundary (using report>area function in fluent). Moreover, when I'm summing up the same faces in fluent (report>sum>mesh>face area magnitude) i get the correct value that coincide with report>area!

So, where's the inconsistency? Am I missing something? Could you suggest me another way to obtain the face area magnitude values I need?
As you could easily guess, i need this to prepare some custom post-processing analysis involving area weighted averages. Being this vital for my work, I thank in advance everyone who helps me.

Regards,

Marco

[Sherlock_1812]

I'm reviving a really old thread because I am facing the same problem now. How to solve the issue addressed in the above question?

[macfly]

Hi,

It is not clear to me which data is exported to Matlab. But if you are running Fluent in parallel mode, maybe it has to do with the mesh partioning, you may find answers in this section and around: http://aerojet.engr.ucdavis.edu/flue...df/node211.htm

[Sherlock_1812]

No this is a serial run with an axisymmetric mesh. While reporting the surface integral along a surface, fluent gives me the right value. When I try to sum up the face area magnitude of each cell on that surface, there is this inconsistency.

I understand, after some reading, that the latter is done through the face area vector projected in the normal direction. Could that be the cause of this?

[nicon97]

Good afternoon.
I cannot tell the right answer because I did not get it entierly. But I can give an advice to calculate more accuratly the result via exporting "Face Area Magnitude" from Fluent.
If you multiply every exported area by the factor 2*PI and sum them, you get a very approximate result that Fluent provide you using surface integral in a given boundary. If the area what you calculate dose not variate in time, we can estimate a correction factor (corr). Its like "corr=AA_1/A_1", where "AA_1" is the area which provide Fluent using surface integral and "A_1" is the sum of all exportes face areas times the factor 2*PI.
So, if you want calculate the heat flux (Q) in a boundary, you can apply this expression: "Q=SUM(qr_i*a_i*2*PI*corr)", where "SUM()" is the summation of all cells of the boundary, "qr_i" is the heat flux per area, "a_i" is the face area and "corr" is the correction factor that I exposed previously. The result is even more near to Fluent calculation, but not exact.
I hope that this can help someone. If there is anyone who can expline how "Face Area Magnitude" is defined exactly, I would be very pleased.
Good bye!

[pakk]

The trick is not to export node values.

[nicon97]

Hello! Thank you very much! It worked. Instead of exporting the default node location values, I exported cell center location values. Multiplying by 2*PI factor the result is exact as Fluent surface integral. Not need in any correction factor "corr" as I defined previously.
Again, thank you very much! Good bye.