Steady state simulation with VOF method

Gottkanzler · May 29, 2017, 15:52

Hi everybody,

I want to simulate the wetting of the evaporator surface inside a falling film evaporator using the Volume of Fluid method. There are two phases in the flow region, a liquid and a gas. The evaporation is currently not considered in the model. The liquid enters a circular channel and flows over the edge of a vertical cylinder. On the inner surface of the cylinder the evaporation takes place.

I know that quad/hexa meshes are recommended for VOF method. I tried creating a block-structured mesh, but could not do it because of the complexity of the geometry. Therefore I generated a tetrahedral mesh including prism elements on the evaporator surface and on the level where the phase interface should be located. My intention for choosing the mesh with prism layers over a simple tetrahedral mesh was to represent the gas-liquid interface better. Can someone confirm this assumption?

The current Model is only running if I choose the transient formulation, but a transient simulation takes way too long. It took days for a few tenth of a second. When I try to run a steady stat simulation it quickly diverges. Does this happen because the initial conditions are far from that that should occur at a steady state?

In this pretty old post (plz help,urgent, vof model steady state ) someone recommended doing a transient simulation with increasing Courant-number to reach a quasi-stationary solution. I tried this, but as soon as I increased the Courant-number over a value of 4 the solution diverges.

Is there a way to reach a steady state solution with the VOF method? If you need any further information about the model please let me know.

I would appreciate it if somebody in this forum could help me resolve this problem.

Diksha · June 14, 2017, 04:17

Hello Gottkanzler,
I am also trying to simulate a steady sate VOF simulation. I am working on condensation in a tube. My problem is that I am not having much condensation. If ever I get the desired condensation, the simulation diverges after some iterations.

Coming to your problem, maybe you can try using the variable time stepping and see the results.

Let me know about the results... if ever you have a solution for me,, please communicate.

Thank you

Gottkanzler · June 14, 2017, 09:32

Thank you for your reply Diksha, I got the steady simulation running now without imediately diverging. Unfortunately I can not help you with your problem. Later on during my Master's Thesis I will have to deal with evaporation in the model, too.

Anyway I am now facing a more specific problem. At some point during the simulation there appear very high velocities at random places in the mesh. The mesh is a pure tetra mesh. The plot shows the maximum velocity in the whole flow region over the iterations. In the other picture there is a volume rendering of the velocity in the flow region. The velocities exceed the maximum expected one by far. What could cause an effect like this?

These are my choices for the Solution Methods:

Pressure-Velocity Coupling: SIMPLE
Gradient: Least Squares Cell Based
Pressure: PRESTO!
Momentum: Second Order Upwind
Volume Fraction: Compressive

The under-relaxation factors do not seem to have much influence on this specific problem.

I am looking forward to every constructive suggestion.

May 29, 2017, 15:52	Steady state simulation with VOF method	#1
Gottkanzler New Member Join Date: Apr 2017 Posts: 4 Rep Power: 9	Hi everybody, I want to simulate the wetting of the evaporator surface inside a falling film evaporator using the Volume of Fluid method. There are two phases in the flow region, a liquid and a gas. The evaporation is currently not considered in the model. The liquid enters a circular channel and flows over the edge of a vertical cylinder. On the inner surface of the cylinder the evaporation takes place. I know that quad/hexa meshes are recommended for VOF method. I tried creating a block-structured mesh, but could not do it because of the complexity of the geometry. Therefore I generated a tetrahedral mesh including prism elements on the evaporator surface and on the level where the phase interface should be located. My intention for choosing the mesh with prism layers over a simple tetrahedral mesh was to represent the gas-liquid interface better. Can someone confirm this assumption? The current Model is only running if I choose the transient formulation, but a transient simulation takes way too long. It took days for a few tenth of a second. When I try to run a steady stat simulation it quickly diverges. Does this happen because the initial conditions are far from that that should occur at a steady state? In this pretty old post (plz help,urgent, vof model steady state ) someone recommended doing a transient simulation with increasing Courant-number to reach a quasi-stationary solution. I tried this, but as soon as I increased the Courant-number over a value of 4 the solution diverges. Is there a way to reach a steady state solution with the VOF method? If you need any further information about the model please let me know. I would appreciate it if somebody in this forum could help me resolve this problem. Ahmed Alkaisi and vats1994 like this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Transient & steady simulation	DIVYA P SOMAN	ANSYS	0	September 3, 2016 15:09
STeady State Simulation	prashantthaker208	FLUENT	0	March 4, 2014 01:48
Steady state VOF problem	Fatima.k	FLUENT	0	February 4, 2014 09:24
anyone done a VOF simulation using steady solver?	R2DM	STAR-CCM+	0	February 24, 2012 15:30
steady state simulation timesteps	amine	CFX	4	March 17, 2008 11:29

June 14, 2017, 04:17		#2
Diksha New Member Diksha Join Date: Mar 2017 Posts: 11 Rep Power: 9	Hello Gottkanzler, I am also trying to simulate a steady sate VOF simulation. I am working on condensation in a tube. My problem is that I am not having much condensation. If ever I get the desired condensation, the simulation diverges after some iterations. Coming to your problem, maybe you can try using the variable time stepping and see the results. Let me know about the results... if ever you have a solution for me,, please communicate. Thank you

June 14, 2017, 09:32		#3
Gottkanzler New Member Join Date: Apr 2017 Posts: 4 Rep Power: 9	Thank you for your reply Diksha, I got the steady simulation running now without imediately diverging. Unfortunately I can not help you with your problem. Later on during my Master's Thesis I will have to deal with evaporation in the model, too. Anyway I am now facing a more specific problem. At some point during the simulation there appear very high velocities at random places in the mesh. The mesh is a pure tetra mesh. The plot shows the maximum velocity in the whole flow region over the iterations. In the other picture there is a volume rendering of the velocity in the flow region. The velocities exceed the maximum expected one by far. What could cause an effect like this? These are my choices for the Solution Methods: Pressure-Velocity Coupling: SIMPLE Gradient: Least Squares Cell Based Pressure: PRESTO! Momentum: Second Order Upwind Volume Fraction: Compressive The under-relaxation factors do not seem to have much influence on this specific problem. I am looking forward to every constructive suggestion.