Nima Khaledy You can't perform a solution just for the turbulence equations without calculating the flow, because the turbulence model is part of the flow equation system. The turbulence model is the closing parameter from the system of flow equations.

If you can't simplify the task to a 2D setting, then I recommend you try to solve the problem in parts. You can solve 1/2 of the calculated volume in one calculation, then get the flow parameters at the border of the calculated volume and use them as boundary conditions for the second calculation, in which you solve the second part of the calculated volume.

Nima Khaledy ,Perhaps I am wrong, but I believe that replacing a two-phase flow with a single-phase one will not reduce the computational cost. In my opinion, it depends only on the number of elements in the grid.

To solve your problem, I would suggest trying to solve your problem in a 2D setting. Although I only use Ansys CFX I assume that the method will be the same and it can be applied to Ansys Fluent.

In Ansys CFX, to solve a 2D problem, you need to create a grid with a thickness of 1 element.

For planar 2D geometries, apply symmetry conditions to the front and back planes. Free Surface Flow Over a Bump is one example of a case that uses this setup. Do not use free slip walls; doing so will hurt accuracy because control volume gradients will not be computed. The extrusion distance should be on the order of the smallest mesh dimension.

For axisymmetric 2D geometries, apply symmetry conditions to the high-theta and low-theta planes unless there is swirl anticipated in the flow, in which case, 1:1 periodic connections should be applied instead. Do not use GGI periodic connections; doing so will hurt accuracy. The extrusion rotation angle for axisymmetric geometries should be small (for example, 1 to 5 degrees)

( link to ansys CFX tutorial (you need watch Free Surface Flow Over a Bump):

https://storage.ansys.com/doc_assets/tutorial_input_files/v190/CFX_Tutorial_Package_r190.zip )

Vladimir Dorofeev ,

Thank you for your answer. I want to study the 3D solution. So, the problem can not be simplified to a 2D model. Also, the geometry is asymmetric.

I have another important question. It may resolve the computational effort. Please check the question that I have written in the following topic:

Solve open channel flow only for turbulence (Ansys-Fluent)?

https://www.researchgate.net/post/Solve_open_channel_flow_only_for_turbulence_Ansys-Fluent

I suggest to see the below link

https://www.youtube.com/watch?v=BnFqgl0oS9w

Regards

Nima Khaledy You can't perform a solution just for the turbulence equations without calculating the flow, because the turbulence model is part of the flow equation system. The turbulence model is the closing parameter from the system of flow equations.

If you can't simplify the task to a 2D setting, then I recommend you try to solve the problem in parts. You can solve 1/2 of the calculated volume in one calculation, then get the flow parameters at the border of the calculated volume and use them as boundary conditions for the second calculation, in which you solve the second part of the calculated volume.

Vladimir Dorofeev

Thank you again for your assistance. I am relatively new to CFD because my field of study is Structural Eng. However, I think there are similarities between Fluid Mechanics and Structural Mechanics. Briefly, they are both involved with solving the governing equations! (often using numerical methods like FEM and FVM)!

Dear Nima,

you touch a delicate problem: a TURBULENT multi-phase open-channel flow. I even assume that you have a high-Reynolds number case. Generally, k-eps is not useful to compute the turbulent viscosity for such topics. Kolmogorov could not solve yet, but have had the better nose and designed HIS version of the two-eq K-Omega model which I then completed here: Simply google it and add my name:

Universal equations and constants of turbulent motion.

Another interesting aspect is the role of non-Newtonicity in multi-phase flows under turbulent conditions and in dispersions etc. - here much depends on the quantitative details. There is a more recent paper and you may google "Baumert Wessling mixing" to find a paper which might interest you about these issues.

You may get me also via [email protected]

Cheers, Helmut

Helmut Ziegfeld Baumert

Thank you for your information.

You are welcome-Helmut

Hi Nima Khaledy

surely it could reduce the time of your simulation, but be careful about your results , because it reduces your accuracy with experimental data about 10% for your validation. But if you want to consider the performance trend of the channel in different condition and won't predict the water depth, I suggest you for using one-single-phase.

I understand that if you omit a phase, it is no longer a multiphase problem. It may be that you save computational resources, but you are dealing with a different system: the conclusions are not applicable to the original system.

Another option (trial and error) is to make the coarser mesh in multiphase, and gradually increase discretization. And compare with single-phase to see if there is a certain convergence between both systems.

Jose Luis Vicens Moltó ,

Muchas gracias. Pero ya no quiero considerar el problema como monofasico. Porque no se observo una reduccion significativa en los costos computacionales. Ademas hay limitaciones en el modelo monofasico.

Hi Nima Khaledy. I don't think it is easy to simplify a multiphase water-air flow problem to one phase, much less in Ansys Fluent. You must take into account that that the VOF model is generally used, which in turn requires a refined mesh mainly near the free surface (which is possible we can hardly intuit where such a free surface is located). If you add a turbulence model to the above (for example k-e RNG, k-w SST), the computational cost increases greatly. I must also tell you that convergence is difficult to achieve in several cases.

I don't know what type of flow problem (open channel, culvert, flow over a weir, stepped channel, flow through a gate, etc.) you are solving, but I can think of the following general recommendations:

1) In ANSYS Fluid:

1.1) Run the simulation with double precision and several processors in parallel (depends on your laptop or similar).

1.2) If the flow can be analyzed as Steady I recommend: model it as Steady; VOF with Implicit formulation; Turbulence with k-e RNG (ensuring that y +> 50, so that you do not resolve the viscous sublayer, otherwise you should to use improved Scalable or Enhanced Wall functions or even the k-w SST model with y +

Leonardo Henao

Muchas gracias!