[flow_CH]

Hello

I have modeled a 3d spillway and its chute for 5 times (different discharges).
The length of model in prototype is 270 meters. The number of cells is 1.6 millions.
The method for multiphase is VOF. All results like depth, velocity, pressure and ... was very good. I defined free surface of flow from volume fraction of water =1 to 0.5.
My problem is about volume fraction 0.5 to 0. Since the mesh of model was large in upper than free surface, the distance from volume fraction of water =0.5 to 0 is very large. (i need the distance from VF=0.5 to VF=0 for void fraction). For better understand please see attached image 1.
I have just 2 months for my thesis and i can not decrease mesh size because the time for iterating is 26 hours for each model. (i have 5 model) and i have extracted results and every thing is OK.

My question is:

For sketching air concentration (void fraction) diagram (attached image 2), can i equivalent VF=0 with VF=0.5 (or c=100 to c=50) since this modeling is true and problem is just from meshing?
I mean can i assume VF=0 to VF=0.5 because my problem is just related to meshing size not for anything else.
I request you and professional researchers to help me please. I need help.

Thanks

[Zaktatir]

I had some students as supervisor and i can say: that 's not Okay to assume that! That will be wrong!

Present your result without any changement and Argue what we are getting and why are u getting this!

[flow_CH]

Originally Posted by Zaktatir

I had some students as supervisor and i can say: that 's not Okay to assume that! That will be wrong!

Present your result without any changement and Argue what we are getting and why are u getting this!

I have to do that because in free surface of water the void fraction is 0.5 and this is wrong. It should be close to 0.9.

[Zaktatir]

free surface is always calculated when VOF is equal to 0,5

[flow_CH]

Originally Posted by Zaktatir

free surface is always calculated when VOF is equal to 0,5

Yes i know. But volume fraction 0.1 should be very close to 0.5 but in my model that it is large because mesh is large. I want the distance from 1 to 0.1 be equal to water depth. Indeed the distance from 0.5 to 0.1 very small that i can assume free surface is in VF=0.1. For this reason i want assume VF=0.5 is equal to 0.1. I hope you understand what i saying.

[Zaktatir]

I understand what you are saying and that is why I am against this assumption because there is no basis for it.

Therefore just interprate your results and argue what you are getting or carry on mesh adaption you have two months and that means you have time to do that

[flow_CH]

It is hard to me for one diagram i spend minimum 1 month. Unfortunately i extracted all results during 2 months and i can not do this again, i have very other works.

[Jim87]

Okay,

First of all I have to support Zaktatir in his is argument. If someone who has experience in CFD is reading "assuming VF 0 as 0,5" you could get in trouble.

It is wrong to assume that VF of zero is the border between two phases in numeric simulation.

But that doesn't mean that your model is wrong or couldn't get fixed.

Some general points:

If you have free surface the Mesh-Quality at the surface has to be fine. If you can use hexahedrons, inflation or even both, that could bring a lot.

In order to get a clear surface (without much merging) your solution method and your viscosity model are very important (name them please)
2D or 3D? If you work easier, do it. 2D is easier to solve and works faster.

- Please name: Fluid, Viscosity and Solver-Model. And a picture of the mesh would be helpfull.

By the way:

The basic problem in dealing with this problems is, that a lot of factors have an influence on a sharp and correct surface.

My experience showed that the initiation and boundaries can influence this as much as mesh and material datas.

[flow_CH]

I think i can not explain this problem good.
I have no.1 diagram in my model. I want to know can i sketch this diagram like no.2 image. (please see Y (vertical axis))

[Zaktatir]

Instead of falsification of the results give a try and carry one as suggested a 2D or adapted 3D simulation.