Learning Forum

[Andrei Ostorukhov]

Hello, for my bachelor these I'm trying to simulate condensation on the surface (phase change from water liquid to water vapor), but after running a calculation I getting an error "floating point". My geometry is basically a cube with a void in the shape of fins. Firstly I have improved my mesh (min. orthogonal is 0.19, average is 0.76; max. skewness is 0.81, average is 0.24; max aspect ratio is 9.23, average is 1.86; min. element quality is 0.21 and average is 0.83), values are now okay, but it still doesn't work, so there is some problem with physics, boundary conditions, or calculation setting. I'm attaching some figures to show them. I'm using a mixture multiphase with phase interaction evaporation - condensation from water to vapor cause everybody does it on YouTube :), when I've tried to set it from vapor to water I got the error "latent heat less than zero". I have also turned on "Surface Tension Force Modeling" and "Surface Tension Coefficient". The primary phase is water liquid and the secondary is water vapor. Boundary conditions. The inside walls (via Fig. problem 1-5 and wall-fluid) have a temperature of 255.15 K and the outside walls (via Fig. Walls) have a temperature of 277.15 K, all in phase mixture. Initialization setting with Fig., after i'm pathing phase vapor by volume fraction with fluid to value 1. 

So what can be the problem?

[Andrei Ostorukhov]

 

 

 

[Rob]

The latent heat error is linked to the to-phase and from-phase for the evaporation/condensation. It's covered in the manual, and the error means you got the phases backwards: Fluent doesn't know what a gas or liquid are so needs to be told. Note, you only need one phase change mechanism as evaporation & condensation are covered in the same one, mass transfer is positive or negative depending on local temperature. 

What time step are you using? That's usually one of the factors in poor/no convergence in these models. 

[Andrei Ostorukhov]

Hello, Rob, thanks for your reply and thanks for explaining this phenomenon with phases, I'm using 0.005 step size. I just came back from uni, where PhD controlled my model and gave me an answer. In my model doesn't work the law of conservation of mass because the vapor in the cube is isolated and pressure is settled to a constant value. Water has way bigger density, during condensation pressure in the system rapidly decreases, leading to error. The model needs a pressure-inlet so it can maintain the pressure. I explained this to someone who will have the same problem and will be looking for a solution.

I tried to set one and now my calculation is running for more than an hour, I hope the result will be close to reality.

[Rob]

Or use ideal gas for the steam. But, yes, whilst Fluent focusses on conserving mass the volume needs somewhere to go too.