Learning Forum

[Jabir]

Hello,

I am trying to simulate the flow of R134a in a multiphase ejector with Lee model for evaporation. The liquid and vapour phases are setup using two different materials created with rgp tables made using the sustemcoupling and refprop. The simulation runs and the flow field seems realistic, though a weired error message is showing up from the initialization stage itself. Please see the images attached. I am using 8 processors and this error appears eight times at cell number 0. But I have checked the total enthalpy and it seems enthalpy of gaseous phase is always greater than that of liquid phase. Then how come latent heat become negative?

In time, printig this error message is getting slower and it slows down the entire simulation. If I increase the number of cores, it become more worse.

Any suggestions are greatly appreciated.

[Amine Ben Hadj Ali]

Ensure that you set the phase change as mentioned in the documentation from "liquid" to "vapor".

[Jabir]

 

 

Could you please explain more? I have set phase-1 as lvapour and phase-2 as liquid. In the phase interactions I setup the phase change as shown below. Please correct me if I am wrong.

 

 

[Amine Ben Hadj Ali]

I think my explanation was fine: phase change to be defined between liquid (from phase) to vapor (to phase). That is mentioned in the Fluent documentation.

[Jabir]

 

I still think your explanation is not clear, thats why I asked again. I belive you are refferring to the Fluent User guide/ Theory guide. Since I cannot go through the entire 5000+ pages today, I have read the section “14.7.5. Evaporation-Condensation Model” in theory guide and section “22.2.10.2.5. Evaporation-Condensation Mechanism” in the user guide. In my understanding what I setup is correct to what explained there.

I have set mixture model with phase-1 as r134a-vapour and phase-2 as r34a-liquid. The interaction between phases is modelled using the evaporation-condensation model. The materials are made using rgp tables created using fluent itself. I am getting an error stating latent heat is negative at cell zero on each computational node.

I believe its a bug in Fluent.

 

[Amine Ben Hadj Ali]

Your phase change definition is wrong: it should be from phase-2 which is liquid to phase-1 which is vapor.

[Amine Ben Hadj Ali]

It is not a bug in Fluent: that is how the Fluent Phase Change Structure is written. Is not my favorite but it is so and as such documented.

[Jabir]

Thanks for your help. I didnt know about such a constraint in Fluent.

[Wiaam Saeed]

So If I wanted to model condensation from vapour to liquid, would I still have to select the phase interaction to go from liquid to vapour and select the "evaporation-condensation" option?

 

[Amine Ben Hadj Ali]

Yes: from liquid to gas. If condensation is happening a negative phase change rate will be reported.