Hello everybody,

I have problems displaying the contours of F_UDMI! When C_UDMIs are used, contours can be displayed. Does anyone know how to display the F_UDMIs? Thank you very much in advance.

Nici Bergroth

F_UDMIs are for faces and they should be the same as the corresponding C_UDMI. I think in the source headers you'll find FLUENT defines these as the same.

FLUENT doesn't save data written to a F_UDMI with the data file as far as I can work out. Basically it is saved in the corresponding C_UDMI or not at all.

Confused? Yeah well I was too. Basically I never really quite worked out exactly what FLUENT does so I stay away from using them.

My guess is that displaying the C_UDMI on the face of interest is equivalent.

Greg

Hi Greg and thank you for your comments! The problem is that when I take sum of the values stored in C_UDMI for the computational volume I get an answer that corresponds to that I've calculted myself and verified as right. These cells are all located adjacent to a wall, because it's the wall that I'm really intereted in. Using the area sum for the wall now for the C_UDMIs, gives not the same answer. Now I belive that the contures displaied for the wall are also not correct!

On the basis of this it would be nice to use the F_UDMI and display the contoures for the wall in the same manner as it is done with the temperature of the wall-outersurface, only for the wall. By being able to do this the sum of the F_UDMIs of the faces of the wall should give the same result as the volume sum for cells. Furthermore, then the contures would be displaied for the wall faces and not for the cell and they would also be correct.

When I'm using only F_UDMIs I can't display anthing, well it show zero for the wall. When doing the same with C_UDMI I get values displaied for the wall.

Well to be honest I'm not sure what you should do.

Perhaps I'll explain what I did, since my problem was in many ways similar.

Basically I wanted to store information only at faces and in some cases in cells and at faces. I found the best thing to do was for each face variable I wanted to store I created a separate UDMI and stored the values in the adjacent cell CV to the wall. I never stored cell and face data with the same UDMI.

Now the face values I was interested in usually relate to a flux - say kg/m2-s across the face. When I stored in the C_UDMI I also converted this to a volume source if that makes sense, based on the cell face area. The reason for doing this is that you can then apply a volume integral in the FLUENT post-processor to get the results. which will then be a total surface flux!

I also define custom field functions to claculate the flux at the wall. These can be used to calculate surface integrals etc.

I'm not sure that its absolutely necessary to store the values as volume sources, but it works well for me. If you store them as fluxes I think you can still do the calcs OK, but just be careful.

Using this approach I have been able to plot contours of the surface values stored in the UDMI.

The other problem you have if you store values in F_UDMI is that they aren't saved with the data file. So if you reread the grid you won't get those values. In my cases this disrupts the iterative process so that's why I store them in the adjacent cell.

As I mentioned before I never really worked out the point of F_UDMIs....anyway these are the work-arounds I've come up with and they work well for me.

Greg

Well, it sound as I'm already doing nearly the same things. I'm working with a DPM model to track particles and calculate their mass flow at a certain wall face. Only problem is that I'm not convinced that the contures ploted now for the wall faces are correct.

For those cells that are located in a corner, 3 faces get the same value if stored in C_UDMI! The F_UDMIs would be very good if they only could be displayed and used for post-prosessing.

Nici

You're right about corner cells - but in my case I have just accepted this. What my model does is add all the values together so that the source terms are correct its just that the displayed plot may not be in that cell.

I agree - I played around with F_UDMIs for a while and then gave up. If you can work it out let me know, but basically I'm happy with the way I've got it setup now, even though its not optimal.

I am pretty sure I get the values reported correctly in the plots. At least I've been assuming this, and they look reasonable.

Porbably you should test this with say a simple known function or single value.

Greg

Well, I'm working on it and if I don't get it running I have to come up with something else. Thank you for you interested and comments.

Nici

Hej Greg and others,

It came to my knowledge that it is not possible to display the F_UDMI's in version 6.1 of FLUENT. Hopefully it is available in the next version...

Nici

Ahhhh....that makes sense....we'll have to wait and see.

I guess that I see the contours of the near wall cell, which in my case store the face values anyway, and hence it works OK.