[lynn]

When i calculate the flow and solidification in continuous casting process by Solidification and Melting module of Fluent. But the results show that the flow is always in the nozzle, as shown in Figure flow.jpg from attachment, and real flow field is shown in Figure right.jpg from attachment. I am totally confused. Does anyone know the reason? Please help me, very thanks.

[raagh77]

Did anyone tried to solve Solidification including pull velocity ?

I tried to include a pull velocity as described in the
http://www.tchpc.tcd.ie/fluent/Tutor...tory/tut21.pdf

after few iterations flow field gets completely absurd.

If i switch off the solidification and pull velocity flow seams to be realistic.

Regards
Raghu

[ambrish4u]

Originally Posted by lynn

When i calculate the flow and solidification in continuous casting process by Solidification and Melting module of Fluent. But the results show that the flow is always in the nozzle, as shown in Figure flow.jpg from attachment, and real flow field is shown in Figure right.jpg from attachment. I am totally confused. Does anyone know the reason? Please help me, very thanks.

I think u r solving with steady state condition...solve it by transient state condition and with proper boundary conditions....see the tutorial.

[raagh77]

I am trying to solve transient case.
As I said flow willl be quite stable until unless Solidification is switched on.

I am trying to use moving wall BC for mold wall regions [moving with casting speed]

@Lynn,
Sorry, if I hijacked your thread!
Hope we both are facing similar problem and we do get some help in the forum here

Regards
Raghu

[lucky_m_m]

hi guys,

Even I am also solving same problem. For transient state I am able to get a converged solution but for steady state I get totally absurd results with most of the time divergence in temperature. I am not able to understand the reason for problem in steady state because the paper that I am trying to validate has solved the problem in steady state.

Could you guys suggest me something as you all have also worked on the same problem.

[raagh77]

@Saurab,

May be with "Solidification" model best approach is to solve in transient (with my limited knowledge I am making this statement).
But even with transient approach the shell thickness developed is not realistic. [2D case was just fine but in 3D the shell thickness and the flow field is complete haywire]

If you are trying to solve thin slab caster solidification [with thickness < 100 mm] then the problem intensifies further.

I am in the process of developing this solidification model with the Ansys customer support and yes hoping to get it solved.

I will update here once I get some meaningful results with Fluent

Regards
Raghu

[lucky_m_m]

Thanks a lot Raghu. Can you give me more references related to it. or could you tell how to use a model equivalent to SIMPLER model in FLuent because the paper I am referring uses SIMPLER model as lot of pressure correction is required in the problem.

[Dr.TDAJones]

This is a really old post but here goes.

I model solidification by first running a steady state with solidification switched off to enable the thermal effects to reach equilibrium.

I then use this result as my initialisation for a transient simulation looking at applying solidification.

If I do it this way I get more realistic results than trying just steady state with solidification, or running the whole thing as transient.

What I am wondering is do I need to apply pull velocities? So far I have not...

[raagh77]

Hi Dr.TDAJones,

"What I am wondering is do I need to apply pull velocities? So far I have not.."

Yes it is better to model walls as moving wall with velocity equal to pull velocity.
Also make sure your heat flux profiles are realistic.

Regards
Raghavendra