[acalado]

Hi gents,

I have a problem where I have a body of a train car that is defined with an "interior surface" and an "exterior surface", basically assuming there is a certain thickness of void space in between.



I know if I had assumed a zero thickness surface I could have shadow surfaces with temperature coupling and a pseudo-thickness.
Is there a way to somehow map/interpolate and associate these exterior and interior surfaces which are offset and different from each other so they behave in the same way as coupled?

Thanks!

[AlexanderZ]

you may try to make a loop over faces of internal wall and store temperature according to coordinates
using other function you may apply stored data as a temperature boundary condition on external wall (with some kind of interpretation, based on coordinate)

[vinerm]

If there is a solid zone between interior and exterior, then there should be a solid body existing between the two surfaces. Or is there nothing between these two? If there is nothing, then either you can do it using UDF, as suggested by Alexander, with a slight modification. Instead of temperature, you need to match the heat flux. Matching the heat flux will lead to conservation of thermal energy. Another option is to create an interface. You can create a translational periodic interface. That will ensure the conservation of fluxes.

[acalado]

Thanks for the suggestions.

Indeed right now there is void space in between (in reality should be solid). I get the idea of using a loop or profile to read the variable of say heat flux, and then apply that through an xyz interpolation to the other face but am not quite sure how to write that.
Do you have any example UDF ?

Thanks!

[vinerm]

An example is not available, at least not with me and it won't be straightforward or easy because Fluent uses polygonal mesh approach to do this, however, here is one process that can be followed

1. Use DEFINE_ADJUST macro along with multiple UDMs to save intermediate data
2. Within the DEFINE_ADJUST, loop over the faces of each boundary, i.e., over interior and exterior. Look at the first example available at the following link on how to do this.

https://www.afs.enea.it/project/nept...df/node100.htm

3. Within each loop, extract the gradients of the field variables. If one of these is solid, then only temperature gradient needs to be extracted. Save this in UDM. Loop over other boundary and use the values from the UDM to extrapolate data. Then repeat the operation backward, i.e., save gradient from second boundary for use at the first one.