I need your help guys, I want to analyse a conjugate heat transfer problem! The Flow region is porous with heat generation in the solid! So I have to solve two energy equations and only a set of N-S-Equations. The question is should I model the Problem as a two Phase Problem (with a non flowing phase) or as one Phase probelm and treat the solid temperature as a user-scalare!! which variant gives me more flexibilities to define the inter-phase heat and momentum transfer (i need hier to implement my own correlations)?? And what about setting the conducting boundray condition for the user scalare is it possible? I have no experience with cfx so every tip from you is a great help for me! Thank you a lot! hanni

Hannibal,

I've done this with CFX-4, but I think it can be done in CFX-5. By far the easiest method is to simulate the solid enthalpy with an additional scalar. There is then a matching source sink at each node to transfer heat between the fluid and the solid based on the surface area, heat x-fer coefficient, and the temperature difference between the two phases. You'll have to divide the solid enthalpy by it's Cp to get it's temperature. This source is best done in USRSRC in CFX-4 but can probably be done through CEL in CFX-5 and applied as a sub-domain source. Also, you'll have to modify the solid and liquid thermal conductivities to account for the porosity.

Hope this helps, Jeff

Hi jeff, thanks a lot for you answer. It was very helpfull!! I looked in the CFX5 solver capabilities and i found that a general form of the transport equation for an Additional Variable (non-reacting scalar) is implemented, in which (I think) there is no way to set the convection term to 0. Can somebody say if i'm wrong with my interpretation?? Is it possible in cfx5 to implement a user-defined (heat conduction) transport equation with a thermal temperature dependent thermal conductivity and waht about the BC?? Thanks and regards hanni

Hi Hannibal,

It is not available from the GUI, but if you edit the CCL, you can change "Transport Equation" to "Diffusive Transport Equation" to remove the advection term.

You can change this in Pre by right clicking on the Domain models object and choosing "Edit in Command Editor". Note that if you re-open the Domain Models panel, your changes will be lost. Alternatively, you could edit your def file to do so.

Regards, Robin

Hannibal,

In addition to Robin's comments, you can set the conductivity (the scalar's diffusivity) to a temperature dependent value through CEL. Simply calculate it as a function of the media's temperature (H/Cp).

For your last question, you may have to use CEL here as well. A fixed wall temperature will need to be converted to an enthalpy value (since the new scalar is the solid enthalpy) as in Cp*Twall.

Hope this helps.

thank you guys! hanni