Learning Forum

[KK]

Hi everyone,

I am modelling the condensation of humid air on a cold wall in a room with a reasonable mesh quality, as shown in the figures below. Hot outdoor air will enter the room with a velocity of 1 m/s and a temperature of 308.15 K. The cold wall has a constant temperature of 289.15 K. The air flow, temperature distribution and humidity levels are the subjects of interest in the study.

 

After conducting some literature review, I am using Multiphase model (Mixture) with Species model (Species Transport), where phase 1 is water-liquid and phase 2 is a mixture-template of water-vapour and air. Phase 2 water-vapour will condensate into phase 1 when the cell is at saturation temperature. Realizable k-epsilon with enhanced wall treatment is used for the viscous model.

The settings for boundary conditions are as follows:

inlet (mixture) - Thermal temperature [K] 308.15

inlet (phase 1) - Momentum velocity magnitude [m/s] 1

inlet (phase 2) - Momentum velocity magnitude [m/s] 1

inlet (phase 2) - Species mass fractions h2o 0.02

inlet (phase 2) - Multiphase volume fraction 1

outlet default

wall 1 (mixture) - Thermal temperature [K] 289.15

wall 2 acts as adiabatic wall

where wall 1 is the cold wall in the middle of the room and wall 2 is the surrounding wall of the room.

Coupled is selected for Pressure-Velocity Coupling where all of the spatial discretization is set to a higher order.

There is no UDF being used.

Before I clicked on initialization, I have checked my case under run calculation and there is no recommendations.

 

The error message "Node XXXXXX: Process XXXX: Received Signal SIGSEGV" occurs when I click on initialization, both hybrid and standard initialization give me the error. Whenever I get this error message, Fluent crashes and it couldn't be opened again until I reboot my computer (all the settings have gone missing and I need to setup again). I also experienced the same issue when I used my university computer. This error shows up every single time when I click on the initialization.

Thanks

[Rob]

If you turn off the multiphase part and run with just air & vapour does the model behave? What RH do you see on the domain assuming it does? 

[KK]

While running only with air and water vapour, there is no mass transfer. The residual plot converges well. The results for velocity and temperature make sense. Water vapour volume will remain constant at the value I initially gave since there is no mass transfer.

When I run multiphase without species transport, it will not give me any errors. The same goes for when I run species transport without multiphase. The error message described above only appears when I run multiphase with species transport.

[Rob]

Which can mean the time step is not suitable for the mass transfer rate, or that the mesh isn't refined enough (or is too fine) to capture the gradients. If you plot RH what does the contour look like? Please post the image here.