[Isaacibp1401]

In the Stefan problem an interface is divided by vapor and saturated liquid. The problem consist on track the interface position Vs time. I solved the problem analytically and also numerically (using finite volume), however now I'm trying to solve the problem using VOF in fluent. For the phase change I'm compiling a UDF, the UDF that I'm compiling is one published by the tutorial of fluent for film boiling. However with fluent and VOF I'm not getting the same rate of evaporation. The rate of evaporation has a similar behavior to the analytic and numerical solutions that I performed but with a lower magnitude.

For the simulation I'm using a 2D geometry (even though the analytic and numerical solution that I performed considers the energy equation just in one direction.) I'm using symmetry boundary conditions on the lateral sides and wall at the top and bottom sides. In the bottom wall a constant temperature is imposed. The dimensions of the geometry are 0.2 m height and 0.05 m width. My initial conditions are a film of vapor at the bottom (0.001 m) and the rest is saturated liquid. The film of vapor has a temperature profile which goes from Twall to Tsat.

Any idea about what could be the problem? One thing that I notice is that the I have velocities with both x and y components, whereas it is supposed that velocities with just one direction should be present (parallel to the heat flux). This might be affecting the evaporation rate. I tried to impose a fixed value and put Vx=0, however when I do this fluent is showing me an error which says "divergence detected in amg solver pressure correction turning off correction scaling".

I also tried to adjust the UDF by for example multiplying the mass source term by a factor of 2.0, and see what happen. However I got similar results, as if I wouldn't have modified the UDF.

Any idea or suggestion will be fully appreciated.

Thanks

[vidyadhar]

Originally Posted by Isaacibp1401

In the Stefan problem an interface is divided by vapor and saturated liquid. The problem consist on track the interface position Vs time. I solved the problem analytically and also numerically (using finite volume), however now I'm trying to solve the problem using VOF in fluent. For the phase change I'm compiling a UDF, the UDF that I'm compiling is one published by the tutorial of fluent for film boiling. However with fluent and VOF I'm not getting the same rate of evaporation. The rate of evaporation has a similar behavior to the analytic and numerical solutions that I performed but with a lower magnitude.

For the simulation I'm using a 2D geometry (even though the analytic and numerical solution that I performed considers the energy equation just in one direction.) I'm using symmetry boundary conditions on the lateral sides and wall at the top and bottom sides. In the bottom wall a constant temperature is imposed. The dimensions of the geometry are 0.2 m height and 0.05 m width. My initial conditions are a film of vapor at the bottom (0.001 m) and the rest is saturated liquid. The film of vapor has a temperature profile which goes from Twall to Tsat.

Any idea about what could be the problem? One thing that I notice is that the I have velocities with both x and y components, whereas it is supposed that velocities with just one direction should be present (parallel to the heat flux). This might be affecting the evaporation rate. I tried to impose a fixed value and put Vx=0, however when I do this fluent is showing me an error which says "divergence detected in amg solver pressure correction turning off correction scaling".

I also tried to adjust the UDF by for example multiplying the mass source term by a factor of 2.0, and see what happen. However I got similar results, as if I wouldn't have modified the UDF.

Any idea or suggestion will be fully appreciated.

Thanks

Hello Isaac,

In the last few weeks, i have also started working on evaporation. I am also at the same stage in which you were as mentioned in the above post:

With almost same setup as yours, entire liquid in my domain instantly evaporating as soon as the simulation has started.
I am attaching the UDF that i have written. I request you to guide me in validating with stefan problem.



Thanks in advance!
vidyadhar

[TaiDangDinh]

Hello I come from Vietnam. Have you mastered this yet? Can I ask you a few things? I will add udf function somewhere in fluent.

Thank you very much!
TaiDangDinh from VN