[marsa27]

Hi, I'm me interested to use a negative "Total Fluid Mass Source" in my VOF simulation to let out air from my mold for the glass forming. I'using a VOF omogeneous model. How I can let out only air from the wall rather than the glass, which must to touch the walls of the mold to cool down without going out too? Is necessary to use a non omogeneous VOF model? In this last case how can I calculate the correct Cd between air and glass? Is it possible set another option to avoid the specification of the Cd?

Thank you

[Gert-Jan]

Can't you apply the Degassing boundary condition? This will dstinguish between air and glass

[marsa27]

Originally Posted by Gert-Jan

Can't you apply the Degassing boundary condition? This will dstinguish between air and glass

No because the walls of the mold are interfaced with the solid. So I cannot use an outlet condition for these interfaces. Moreover the degassing condition requires a fluid dispersed (air dispersed in the glass) and accoriding to me it is not very correct, because I should set the diameter of fluid dispersed and the cd for the not omogenous model. Am I wrong?

[ghorrocks]

How does the air and glass exist? Is the air mixed in the glass as bubbles or inclusions? Is it dissolved in the glass? Or are there clear boundaries between glass and air?

[marsa27]

Originally Posted by ghorrocks

How does the air and glass exist? Is the air mixed in the glass as bubbles or inclusions? Is it dissolved in the glass? Or are there clear boundaries between glass and air?

There are clear boundary between air and glass as in the Figure attached. In fact this simulation is the final forming of the glass parison in a final mold trough a air blowing form the top. Instead the air between the parison and the mold must be expeller during this blowing

[ghorrocks]

In that case you should use a homogenous multiphase model with a free surface model. Have a look at the air/water tutorial example of "free surface flow over a bump" for how to set this up.

This should also make the task of working out how to expel the air easier - you just do whatever is done in the real mould to expel the air. It probably has some venting ports somewhere. So just include the vent ports as normal pressure openings or outlets and you won't have to do any special treatment with them. That is, your gas only mass sink is not required.

[marsa27]

Originally Posted by ghorrocks

In that case you should use a homogenous multiphase model with a free surface model. Have a look at the air/water tutorial example of "free surface flow over a bump" for how to set this up.

This should also make the task of working out how to expel the air easier - you just do whatever is done in the real mould to expel the air. It probably has some venting ports somewhere. So just include the vent ports as normal pressure openings or outlets and you won't have to do any special treatment with them. That is, your gas only mass sink is not required.

Unfortunately in the reality the mold is not completely hermetic, while in CFD the domains are closed. I'have already tried to use some patchs to put fictitious outlets, however when the air blow in pressure from this outlets escapes also the glass. For this reason I am intereseted to use a condition wall only for the glass and outlet for air. How can I do?

[ghorrocks]

Does the air escape on the mould parting line, does the mould have lots of little air vent holes in it, or is the mould porous to air?

[marsa27]

Originally Posted by ghorrocks

Does the air escape on the mould parting line, does the mould have lots of little air vent holes in it, or is the mould porous to air?

Exactly, but how I can simulate this situation to let to the air to escape, without that also the glass escapes?

[ghorrocks]

I was asking a question, not making a statement. Which of those three air venting options is it using?

[marsa27]

Originally Posted by ghorrocks

I was asking a question, not making a statement. Which of those three air venting options is it using?

Initially I have put some holes at the opposite position of the glass while the air blowing enters from top, but after some times also the glass escapes from these when is compressed (this problem I could resolve to adjust the time duration). However in the first part of the mold, some air remains trapped, so the glass doesn' t touch the mold without colding; instead in the bottom escpaes air ad then also glass. For this reason according to me colud be correct put a conditions for all the walls of the mold as wall only for air and outlet for the air. How can I do?

[ghorrocks]

I did not understand your last post.

From my knowledge of glass blowing, they put either vent ports in the mould or they use the parting plane as a vent. Is this what is happening in your case? Or is it using something else?

[marsa27]

Originally Posted by ghorrocks

I did not understand your last post.

From my knowledge of glass blowing, they put either vent ports in the mould or they use the parting plane as a vent. Is this what is happening in your case? Or is it using something else?

Yes I use some outlets with a pressure outlet, but also the glass exits from the domain. So to avoid this, I would set a wall condition only for the glass at all the walls of the mold

[ghorrocks]

Let me put this bluntly - you are proposing a poor fix for a problem you should not have because you are modelling the entire thing wrong in the first place. Your questions is a classic X-Y problem (https://en.wikipedia.org/wiki/XY_problem)

You need to be using the correct multiphase model before you consider how to handle the air. You need to be using a homogenous free surface multiphase model, not a dispersed model.

[marsa27]

Originally Posted by ghorrocks

Let me put this bluntly - you are proposing a poor fix for a problem you should not have because you are modelling the entire thing wrong in the first place. Your questions is a classic X-Y problem (https://en.wikipedia.org/wiki/XY_problem)

You need to be using the correct multiphase model before you consider how to handle the air. You need to be using a homogenous free surface multiphase model, not a dispersed model.

In fact in the third post of this thread I have written that a dispersed fluid is wrong.. It was another person who offered me the degassing condition and I said that in my opinion it is wrong. I am using a VOF homogeneous model and my problem is that I would that only the air exits from the mold without the glass. I think I was clear. Do you know the solution for my problem? My questions are those reported in my first post... then I have attached a figure in answer to you in the 5th post that didn't understand why the air exists in my problem

[ghorrocks]

I see you did say that in your first post. Sorry about missing that, looks like I overlooked that comment. My apologies.

Some ideas:
* You could use vents with normal openings in them, but at the start of the vent use a momentum source term to stop the flow with a bulk source making it only apply to the glass volume fraction.
* Quoting the documentation, Solver Modelling guide, section 7.16.3 "The phasic mass flow outlet condition is not available for homogeneous multiphase simulations" - so you can't make an outlet which does not draw a specified phase.

So at first glance it looks like the bulk source term acting on a momentum source term is the only one I can think of at the moment. You may be able to find other ways if you hunt around.

But if the vents are small in diameter, then won't the very high viscosity of the glass mean that it will not travel far into the vent? Then you don't need any special treatment.

[marsa27]

Originally Posted by ghorrocks

I see you did say that in your first post. Sorry about missing that, looks like I overlooked that comment. My apologies.

Some ideas:
* You could use vents with normal openings in them, but at the start of the vent use a momentum source term to stop the flow with a bulk source making it only apply to the glass volume fraction.
* Quoting the documentation, Solver Modelling guide, section 7.16.3 "The phasic mass flow outlet condition is not available for homogeneous multiphase simulations" - so you can't make an outlet which does not draw a specified phase.

So at first glance it looks like the bulk source term acting on a momentum source term is the only one I can think of at the moment. You may be able to find other ways if you hunt around.

But if the vents are small in diameter, then won't the very high viscosity of the glass mean that it will not travel far into the vent? Then you don't need any special treatment.

Thank you for the answers.

*As bulk source what what do you intend? Beacuse I see that there are three equation soruces: for energy, for eddy dissipation and for TKE. Instead a source for the continuity equation setting a negative massflow rate (sink) only for the air I already tried with a homogeneous model, but in reality also the glass escapes out.

*According to you, using a not homogeneous multiphase model, if in the outlet I set fluid dependent and then in the fluid values tab I set for the glass a zero massflow rate, while for the air a massflow rate>0, is correct?

[ghorrocks]

Bulk sources can be applied to a momentum equation, and that is what I suggested.

You cannot define a mass source/sink to only gobble up one phase (as you have found out). There is no way I know to change this.

inhomogenous model question - I don't think so. But feel free to have a look in the documentation and try some ideas.

As I said, the only way I can think of doing this is to put a normal pressure opening in a small port and use a bulk source term on the momentum equation in the port to stop the glass phase going down the port (so the glass never gets to the opening).

[marsa27]

Originally Posted by ghorrocks

Bulk sources can be applied to a momentum equation, and that is what I suggested.

You cannot define a mass source/sink to only gobble up one phase (as you have found out). There is no way I know to change this.

inhomogenous model question - I don't think so. But feel free to have a look in the documentation and try some ideas.

As I said, the only way I can think of doing this is to put a normal pressure opening in a small port and use a bulk source term on the momentum equation in the port to stop the glass phase going down the port (so the glass never gets to the opening).

Unfortunately I don't better understand how setting a bulk source for the momentum equation and what to specify.
I have set in an small patch an opening pressure condition and then in its source tab, after to have activated "bulk sources" I cannot to find the momentum equation. I attach the options that I can see with a figure.

Am I doing something wrong?

[ghorrocks]

The opening is just a normal pressure opening. No special treatment at all. But in the port leading to the opening you apply a volume bulk momentum source to stop the glass phase only.