Evaporation-Condensation in a porous zone

  • Last Post 26 July 2018
maximilan posted this 20 July 2018

Hi there,

I'm in the midst of my bachelor's thesis and I am supposed to model a moistened insulation.

The insulation material is supposed to be foam glas gravel, which I modeled as a porous zone inside a cube with 1x1x1m as dimensions. The cube is modeled as a cube with walls on all sides. There is no in- and outflux.

My goal is to simulate the heat transfer from the heated bottom (fixed temperature) to the cooled top (fixed temperature) of the cube, in order to get the heat flux through the insulation. The heat flux is monitored through the heat flux trough heated bottom wall of the cube. With this information it should be possible to calculate the effective heat transfer coefficient.

The effect of evaporation-condensation of water inside the insulation on the heat transfer is tried to be found. I am using the Mixture-Model with three phases: air, water-liquid and water-vapour. I activated the Evaporation-Condensation Model. I patched the volume with air and water-vapour with a volume fraction of saturated air at the patched temperature of 308.15K.

I do get phase changes inside the cube with the evaporation-condensation model. The standard state enthalpy of water-liquid and water-vapour are changed, so that they fit the actual latent heat of vaporization (4.40e+07 J/kmol). In my tries I also tried different standard state enthapy values for both water-liquid and water-vapour, but the heat transfer didn't change. 


My main problems are:

- As soon as I try to simulate in a steady simulation I get temperature values above 5000K and below 1K, and Fluent returns a floating point exception.

- If I simulate the model in a transient simulation and let it run until the transferred heat through the bottom of the cube is constant, the transferred heat is way to high. (According to measurements it should be around 7 watts, but I'm getting something like 30-40 watts).


Has anyone ever tried to model a closed volume with the evaporation model and got it to work? Has anyone got any idea why the heat flux is so high?

Any help would be highly appreciated, as I am on this simulation for 3 months already and starting to get pretty frustrated.

Thanks! Maximilian

Order By: Standard | Newest | Votes
abenhadj posted this 20 July 2018


Can you verify if the overall heat balance is fulfilled? Which material laws are you using for the cavity? Can you check the mass of air whether it remains constant or not?


Best regards, Amine

maximilan posted this 20 July 2018


thank you for the quick reply.

The overall heat balance is fulfilled!

The cavitation model used is: Singhal et al.

The mass of air remains constant.


Best regards,


abenhadj posted this 20 July 2018


Why are you using a cavitation model: the mass transfer is driver by thermal effects. 



Best regards, Amine

rwoolhou posted this 20 July 2018

To add, you need two phases:  water liquid and a air - water vapour species mixture. That should simplify things a bit too. 

maximilan posted this 25 July 2018

Hey there,

I checked again and cavitation is turned off.

I added the air-vapour species mixture. The simulation is still running, but it seems like the heat flux is still to high.

There is one more thing that I don't get. I want to calculate the heat flux that is reached in a steady state. If I turn my simulation into a steady simulation I do get a floating point exception immediately. My workaround is to create a transient simulation and wait until the heat flux doesn't change anymore. 

Has anyone got a idea if there is any other way to set it up, so I can run it steady?

Thanks for all your replys!

abenhadj posted this 26 July 2018


I will stick to transient analysis which would offers robustness compared to a steady-state run especially if the flow at the beginning and before it settles itself down is not steady at all.

Please use a pressured based saturation temperature and not a constant saturation temperature as I assume that you will have a sort of pressure buildup in your closed domain (and that is why not steady-state).





Best regards, Amine