[Dragonuser]

Hello, everyone!
I am a new user here, but I have long visited and studied from here after I laid my hand in CFD.
Here is the recent problem I met when I was simulating the steam condensation.
The model used is very simple. The main pipe contains superheated steam and the sub-pipe contains sub-cooled water. When they are mixed together, there must be condensation and the flow regime is supposed to be stratified flow. So I use homogeneous model and phase change. The property of the material is IPAWS IF97 for both water and steam. I have extended the range of pressure and temperature in case it exceeds the limits.
Still , I met something like this:
+--------------------------------------------------------------------+
| ****** Warning ****** |
| |
| Independent variables were clipped during table generation |
| at: END OF TIME STEP |
| 0 detailed warnings were printed because the maximum |
| number of warning messages was exceeded. |
| |
| The maximum number of detailed table clip warning messages can be |
| controlled with the following expert parameter: |
| |
| max table messages (default: 0) |
| |
| Please increase the number, if you need further details of the |
| locations and variables involved. |
+--------------------------------------------------------------------+



And after about 90 interations,
I met
ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Floating point exception: Overflow |
| |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Stopped in routine FPX: C_FPX_HANDLER |
| |
| |
| |
| |
| |
I suspect of IPAWS IF97, but it seems all right since so many people have used it. So I am really confused and don’t know what to do.
BTW, I am using ANSYS 15.0.

[ghorrocks]

The model used is very simple.

Are you kidding me? A multiphase simulation with phase change? Using IAPWS material properties? This is not a simple model at all.

The overflow error is an FAQ: http://www.cfd-online.com/Wiki/Ansys...do_about_it.3F

The table clipping error is probably due to the divergence in the overflow error.

And you should not limit the table clipping error messages as then you have no idea what is going on.

[Dragonuser]

Thanks for replying so soon, I was really looking forward to solve this.
And by ‘simple model’, I mean the geometry model, sorry for the misunderstanding .
And I have read some of the previous questions about IPAWS. And I want to find the “max table messages’ and extend it to a larger number. But I look it everywhere in the ‘expert parameter’, really have not a clue where it is.
I really have not come up with any idea how to deal with this. So would you give me something more specific?
Thanks anyway.

[ghorrocks]

The table generation settings is under the material properties tab.

[Dragonuser]

I am not sure if I have made it clear. But by
| max table messages (default: 0) |

I mean how to set this in expert parameters

not 'table generation settings '.

And I have double checked table generation settings ,there
is no such thing as 'max table messages'.

[ghorrocks]

I do not know where the maximum table error messages setting is coming from. But it is not vital anyway, your key problem is numerical instability. The numerical instability (ie the overflow error) is probably causing the table error, not the other way around. So when you fix the numerical instability you will probably fix the table error.

[Dragonuser]

So is it okay to say that this problem may not result from the ipaws,

cuz my classmates and I always assume there is something inherently wrong in ipaws water. ie, when using constant property water, we can get a convergent result whereas ipaws water cannot.

And if so, I will focus more on the other settings, not the property of water anymore.

Thanks again in advance here.

[ghorrocks]

There is nothing wrong in IAPWS. It is a highly verified and very accurate material model. The problem is that IAPWS is a non-linear material model and is therefore more difficult to converge than a simple constant properties material model.

[Dragonuser]

Again, thanks a thousand times.

I still cannot understand what might be wrong here.

I am now using outlet boundary for two-phase steam and water. Every now and then, I receive something like

+--------------------------------------------------------------------+
| ****** Notice ****** |
| A wall has been placed at portion(s) of an OUTLET |
| boundary condition (at 12.4% of the faces, 9.2% of the area) |
| to prevent fluid from flowing into the domain. |
| The boundary condition name is: outlet. |
| The fluid name is: gas. |
| If this situation persists, consider switching |
| to an Opening type boundary condition instead. |
+--------------------------------------------------------------------+
+--------------------------------------------------------------------+
| ****** Notice ****** |
| A wall has been placed at portion(s) of an OUTLET |
| boundary condition (at 12.4% of the faces, 9.2% of the area) |
| to prevent fluid from flowing into the domain. |
| The boundary condition name is: outlet. |
| The fluid name is: fluid. |
| If this situation persists, consider switching |
| to an Opening type boundary condition instead. |
+--------------------------------------------------------------------+

I vaguely recall that someone recommended that for two-phase flow, it is better to set up the opening instead of outlet boundary.


But actually, you need to define the temperature and void fraction at the opening. So here comes the question: how can you define something you want to solve or you do not know beforehand. And if it is incorrect, then can you get convergence?

Again, hoping to hear from you, Mr. Glenn.

[ghorrocks]

The message is a warning. If it goes away after a few iterations then you can ignore it. If it stays then you better do something about it - which usually means shifting the outlet boundary further downstream to be out of the recirculation. If you do this (rather than defining it as an opening) then you do not need to define backflow conditions as there is no backflow.