Learning Forum

[Deepak_296]

I want to simulate bubble formation in a container partially(3/4th) filled with water. Gas injection through a hole in the bottom. This needs to be a 2D case. Can I do Multiphase ---> Eulerian----> Multi-Fluid VOF model?

[Rob]

You can. You probably want sharp/dispersed but I'll let you read up for why.

[Deepak_296]

What is the equivalent of Max face size in Ansys 2020 R1 meshing? This option isn't there in 2020 R1. I'm following a tutorial but it's for an older version

[Rob]

Can't remember without looking, if you look in the Sizing section it's relatively straight forward to switch from old to new.

[Deepak_296]

Thanks Rob.
For the bubble formation simulation in water, I need to have a gas phase at the top of the water with an interface separating the two phases. I can say the container is 3/4th filled with water. Could you please suggest to me how to get that?

[Rob]

If you mean free surface between the phases look at the Patch function in Fluent. You may also want adaption registers to help. Note, "interface" is often used to specify a change in cell zone so be careful with how you word questions: engineers are nearly all pedants! ;)

[Deepak_296]

I meant to have a container 80% filled with water having air at the top. I managed to get that using adaptation registers followed by patch to defined the volume fraction of water. However, when I am running the case, it is throwing some "floating point exception error." I am using a 100mm X 80mm water tank with water up to 80mm in height. Air inlet velocity is 0.2m/s from the bottom. Coupled VOF model is used with the k-epsilon turbulence model. The Time step size is 5e-4 and the number of time steps is 1000. Iteration=20/time step size. I am clueless why is it throwing the error. Please help me to fix this.

[Rob]

With VOF you're fully resolving the bubbles, is your mesh fine enough? How long does it take the flow to cross one cell?

[Deepak_296]

Yes, the mesh is fine enough. Yet I have not calculated the time for the flow to cross one cell. Am I going well with the VOF model? Or do I need to try some other method?

[Rob]

Re the time, work that out, it wants to be about 1/10 to 1/3 of the time step. Re the multiphase model, it depends: read up on the options then think what you want from the results. That's how you decide which model to use.

[Deepak_296]

I want to track the formation of the bubble and their change in shape and size with time. Also the velocity profile of the water around the bubbles. Somewhat I am getting the result but not accurately as bubbles are getting bigger and bigger due to coalescence. Is it possible to define the initial bubble size to be formed due to gas injection?

[Rob]

OK, so you will need VOF. Bubbles can be patched into the domain (covered under initialisation). You will need a very good quality and well resolved mesh, small time steps and time.

[Deepak_296]

If we patch the bubbles into the domain, this implies that the bubbles are already present in the domain. But In my case bubbles should form as a result of the gas injection from the nozzle at the bottom. Is there any way to fix the size of the bubble to be formed due to gas injection only? And how to measure the diameter of the bubble after its formation?

[Rob]

If you want to set the inlet bubble size you can't use VOF: please read how the various multiphase models work.

[Deepak_296]

I am using Multiphase---->Eulerian----->Multi-Fluid VOF model. I am able to model bubble formation, coalescence, breaking into smaller bubbles, and bursting at the top of the water surface. But I didn't find any option to set inlet bubble size. Is it possible in this model?

[Rob]

No, you can set the bubble size in the phase panel. You can also use population balance. However, as you want to see the bubbles growing etc multifluid VOF or VOF are the models to look at.

[Deepak_296]

Thanks Rob for this information. But I am unable to find that in the phase panel. If you can give me some details on how to set the initial bubble size in the phase panel, it would be very much helpful for me.

[Rob]

It's one of the secondary phase properties.

[Deepak_296]

I tried setting the initial bubble size, but it is giving the same result. Although it is of 1e-5 order, the diameter of the bubble forming first is approximately 1 cm. How is that possible? Am I missing something here?

[Rob]

The bubbles size in the panel is for the Eulerian model. Depending on the flow field and boundary condition you can get anything from a dispersed flow to a stable free surface.

[Deepak_296]

My simulation is giving weird results. The bubble is taking an unexpected shape after the coalescence which is not at all desirable. It should form fine bubbles having almost spherical in shape. i don't know what the possible reason is. Please help me to rectify this.

[Rob]

That looks pretty good: bubbles tend to deform depending on the velocity, viscosity and surface tension effects.

[Deepak_296]

Yes, bubble deformation will occur depending on the above-mentioned parameter. But the bubble is almost 50 mm in size as the total width of the domain is only 80 mm. How can I get smaller bubbles which are confirmed by the experiments? And how to get plume region?

[Rob]

Check surface tension etc, is the model 2d or 3d?

[Deepak_96]

[Deepak_296]

It is a 2D model and I have used the standard surface tension between water and air i.e. 0.072 N/m.

[Rob]

Which is a 1m deep channel. Is the experiment in a round pipe?

[Deepak_296]

No, it is 2D planar (rectangular) model. Dimension is 100 mm X 80 mm. Water is filled upto 80 mm height only and rest 20 mm at the top is air phase. Will the surface tension be different for the dimension I have used?

[Rob]

Read what 2d is defined as in Fluent. How thick is the experimental system?

[Deepak_296]

Yes, the experiment is in a cuboidal vessel. Should that make a difference if I am simulating for 2D model?

[Rob]

What are the dimensions of the experiment? Ie how "thick" is the domain?