Learning Forum

[tumulpurwar]

I want know do latest version of Ansys support polyhexcore, polyhedral advance meshes for multiphase simulation?

As I see about multiphase solver it accepts tetrahedral Or hexahedral Or combination of both.

Is there is some advancement in new version regards to this solver, where we can use polyhedral, polyhedron type advance cells?

[Murari Iyengar]

Hi tumi Multiphase solvers in Fluent accepts all types of meshes. You can create Polyhexcore/polyhedral meshes using Fluent meshing. When launching standalone Fluent, open it in Meshing mode instead of Solution. You can import geometry there and create a mesh. I recommend using the latest version 2022R1 but Fluent Meshing is supported in older versions as well.
For more information on Fluent meshing, you can watch the following video:

[DrAmine]

There is no mesh topology restriction in Ansys Fluent Multiphase Modeling (or any other models).

[tumulpurwar]

I read about limitations from Ansys guide book about limitation

[tumulpurwar]

[Rob]

We tend to avoid poly-hexcore around the VOF interface, but the solver does work on those types of cell.
Now the image loaded...... CLS is an older model and struggles with several newer features. I don't use it.

[tumulpurwar]

Hi Rob
These below statements were given by you few weeks back in one of my post related vof:
Velocity inlet: open channel bc.
Pressure outlet: open channel bc
Sides were symmetry.
Seabed was wall
Sky was a symmetry plane.

I want to know,have you wrote it wrong, as Velocity inlet:open channel doesnot work in VOF...pressure based inlet & outlet work only.Can you explain please?

Thanks

[Rob]

You need to turn waves on (not sure why) and then set the wave option to none.

[tumulpurwar]

Thanks Rob,i was selecting ever "open channel" only instead of "open channel wave BC".thats why was not getting velocity inlet option.
One more question i have, such kind of error i am getting...what can be the reason of such error
The mesh below

[tumulpurwar]

Hi Rob
For your last comment
"You need to turn waves on (not sure why) and then set the wave option to none".
IT MEANS WATER WILL MOVE WITH NO WAVES, JUST WITH CONSTANT DEPTH, AND WILL PASS CROSSING THE SHIP HULL WHICH IS AT STATIONARY POSITION?
i DONT KNOW, IS IT POSSIBLE TO HAVE NO WAVES(I MEAN CREST AND TROUGH) FOR MOVING WATER?

[tumulpurwar]

Hi Rob
If i am trying STEADY STATE , than my solution convergere, but i can understand that results graphically are wrong.
If i do transient state, my graphical results looks correct, but solution only converge on steps, but not globally even after running 3000 iterations.
What can be the reason?
I have further currently increased my iteration to 6000, to see if transient solution will converge or not.

Thanks

[DrAmine]

What do you mean it converges on steps?

[tumulpurwar]

Hi DrAmine, i meant to say this, regarding convergence....and for globally i meants say like i am running for 6000 iteration, it should somewhere stop before 6000, and should tell the solution is converged.

[DrAmine]

Why shall it stop? Are you defining interruption criteria? If you run for 6000 Time Steps it will run all these time steps ( if there is nothing defined to stop the run).

[tumulpurwar]

Hi, I am bit unclear, like in steady state, it stop in between and say solution is converged.
In transient which currently I am doing, don't it ever stop and say solution is converged?

[DrAmine]

I does stop if the requested time steps are finished (or total flow time is reached), or in case you define an interruption criteria it will stop or if you stop. Every time step you might see a time step convergence and Fluent will write a message telling that the time step has converged.

[tumulpurwar]

Ok.. Thanks DrAmine.
Two more question, I wanna ask
What is difference between
Pressure inlet, pressure outlet
And
Velocity inlet, pressure outlet in Multiphase open channel?

2nd question, if my Y* is coming between 3 to 9..and my turbulence model is KW SST... Will it able to give correct results? As we are in buffer layer.

Thankyou

Regards
Tumi

[DrAmine]

Pressure inlet to prescribe the total pressure inlet (including hydrostatic pressure). Velocity inlet is similar to single phase but if you add open channel wave you can account for segregated velocity inputs. Documentation provides more insights.

SST model does habmve mesh insensitive wall treatment so your yplus should be okay if you are putting enough cells in the Boundary layer.

[tumulpurwar]

Thanks DrAmine
Wherever i have checked any research paper regarding calm water test of ship with cfd..they always used Velocity inlet...i also wanted to use velocity inlet, but i dont see option in my license which ROB suggested few days back in his comment in this Post.
This red box i dont have

Note: I dont need waves because its calm water test

[tumulpurwar]

Hello DrAmine
One more question

Why i am not seeing kelvin wave pattern on back side of ship(i took currently for simplicity, ship shape as simple primitive cuboid) with some draft say 0.28 , it is 7 meter in length
my mesh

[DrAmine]

I do not understand the issue with Velocity inlet.

[DrAmine]

Temporal and Spatial resolution whenever you are "playing" with waves especially if the waves appear due to Coriolis forces (not present in your case) or due to bow-waves past the ship.

[tumulpurwar]

Dear DrAmine
To make you understand in other words,for example, if for my ship calm water test simulation, will there be difference in Results,if i once use Velocity inlet based VOf, and one time Pressure inlet based VOF?(as in both cases the inputs which we give in inlet are same, velocity,draft, and reference line)

Quoting your statement "the waves appear due to Coriolis forces (not present in your case)"...why?, because my ship is stationary?

[DrAmine]

We also did recently an analysis for Wigley Ship hull with steady state solver and the results were fine when it comes to drag prediction as well as overall wave patterns.

[tumulpurwar]

May be my mesh is not good that can be possibility , i am refining it further, but please can you answer my above doubts...i ever see in youtube people for hull simulation used Pressure based inlet..but whenever i looked for research paper they always used Velocity inlet based...i dont understand, for me it looks both are same condition...Am i correct in understanding or i am misunderstanding these both inlet conditions?

Also in steady case my solution fails and says its finding difficulty in converging the solution, and if i try Transient, than it converge, what can be the reason of it?i am just doing calm water test of ship model

[tumulpurwar]

Hi DrAmine
Thanks...i had bit misunderstanding..i have that option in velocity inlet, just saw, so now issue is solved.thanks for help

[DrAmine]

You can use whatever you want to use for BC at inlet: only what you now is important. So if you know the speed and free surface level / or the wave-theory just use Velocity BC.

[tumulpurwar]

thankyou got your point..thanks

[tumulpurwar]

Hi DrAmine

I dont understand, why my simulation of vof is failing in STEADY STATE?..but PASSING/CONVERGING in TRANSIENT...why?

[tumulpurwar]

Hi DrAmine
There asre so many types of wall functions under WaLL TREATMENT CATEGORY FOR MY HULL DRAG CALCULATION..FOR K E MODEL, WHICH WALL FUNCTION I NEED TO USE
Also i manually put theses values to 0.1 & 0.01 for the red boxes...normally what are the correct initial value i should give to start with?

[tumulpurwar]

IS ANSYS FLUENT MESHING IS SLOW ?
SINCE 1 HOUR STUCK ON 50% FOR VOLUME MESH

[DrAmine]

Actually not. It is known to be very quick.