[johnsonw]

Hi
would someone could tell me the difference between Axisymmetric and Axisymmetric Swirl? How does the solver calculates the pressure force and mass of rigid body in these 2D case? To do the calculation, you need know specific area and volume.
Thanks in advance

[azna]

Hi ,

I was wondering that did you find out the difference between Axisymmetric and Axisymmetric Swirl? I'm also looking for the answer of this question

thanks

Originally Posted by johnsonw

Hi
would someone could tell me the difference between Axisymmetric and Axisymmetric Swirl? How does the solver calculates the pressure force and mass of rigid body in these 2D case? To do the calculation, you need know specific area and volume.
Thanks in advance

[LuckyTran]

Originally Posted by azna

Hi ,

I was wondering that did you find out the difference between Axisymmetric and Axisymmetric Swirl? I'm also looking for the answer of this question

thanks

The axisymmetric solver does not permit an out of plane velocity component. The swirl solver does.

[hitzhwan]

Originally Posted by LuckyTran

The axisymmetric solver does not permit an out of plane velocity component. The swirl solver does.

what does the plane velocity mean？

[LuckyTran]

Originally Posted by hitzhwan

what does the plane velocity mean？

The axissymmetric solver limits the velocity to the x-y plane (the velocity is 2D). The component of velocity in the z direction is 0. The swirl solver does not and allows for a 3D velocity vector even in a 2D domain.

[rezendervp]

This is explained in the Fluent Solver and Theory Manual.

It is a 2d Axissimetric that the 3rd component of velocity is accounted for in the calculations. That is, you flow is rotating such a burner premixed.

https://www.afs.enea.it/project/nept.../th/node19.htm

[hitzhwan]

Originally Posted by rezendervp

This is explained in the Fluent Solver and Theory Manual.

It is a 2d Axissimetric that the 3rd component of velocity is accounted for in the calculations. That is, you flow is rotating such a burner premixed.

https://www.afs.enea.it/project/nept.../th/node19.htm

Hi， I have a round tube, the flow inside, I want to simulate the flow process, if I use a 2D model, which should I use? Thank you.

[rezendervp]

In theory, you can use both. If a swirling behavior is present (theta-velocity component), this option will deal with the 3rd velocity component in a 2d approach. If this component is null, it will calculate as well, but is a loss of computational efforts to consider a governing equation for a null variable. But it will work.

Usually, in pipe flows, this is not the case, we have radial and axial components at the inlet, and after some length, with a developed profile ( laminar or turbulent) only an axial component. Hence, the simple and common approach is to use the axisymmetric solver. There is no swirl.

[hitzhwan]

Originally Posted by rezendervp

In theory, you can use both. If a swirling behavior is present (theta-velocity component), this option will deal with the 3rd velocity component in a 2d approach. If this component is null, it will calculate as well, but is a loss of computational efforts to consider a governing equation for a null variable. But it will work.

Usually, in pipe flows, this is not the case, we have radial and axial components at the inlet, and after some length, with a developed profile ( laminar or turbulent) only an axial component. Hence, the simple and common approach is to use the axisymmetric solver. There is no swirl.

Thank you, could you tell me an example that has the swirling behavior?

[rezendervp]

It is when the flow has some rotation:





https://www.researchgate.net/profile...cated_W640.jpg

[hitzhwan]

Originally Posted by rezendervp

It is when the flow has some rotation:





https://www.researchgate.net/profile...cated_W640.jpg

You mean one rotating impeller is in the pipe?

[rezendervp]

Yes, it is one possible case, but the impeller is not considered as part of the domain. But an impeller is not needed. It can have some tangential inlet, baffles, or any other thing that may cause the fluid to flow with rotation. Under this scenario, we intend to avoid a 3D simulation, and one may choose a 2d axisymmetric formulation, but considering the rotational or tangential 3rd component of velocity. There is one more governing equation to that component to be solved, but into a 2d domain. It is a smart simplification to save computational efforts and time.

[hitzhwan]

Originally Posted by rezendervp

Yes, it is one possible case, but the impeller is not considered as part of the domain. But an impeller is not needed. It can have some tangential inlet, baffles, or any other thing that may cause the fluid to flow with rotation. Under this scenario, we intend to avoid a 3D simulation, and one may choose a 2d axisymmetric formulation, but considering the rotational or tangential 3rd component of velocity. There is one more governing equation to that component to be solved, but into a 2d domain. It is a smart simplification to save computational efforts and time.

Thank you so much, you give me a detail response. In addition, do you know the udf ？ I have some questions about it.

[yylaw]

Originally Posted by rezendervp

In theory, you can use both. If a swirling behavior is present (theta-velocity component), this option will deal with the 3rd velocity component in a 2d approach. If this component is null, it will calculate as well, but is a loss of computational efforts to consider a governing equation for a null variable. But it will work.

Usually, in pipe flows, this is not the case, we have radial and axial components at the inlet, and after some length, with a developed profile ( laminar or turbulent) only an axial component. Hence, the simple and common approach is to use the axisymmetric solver. There is no swirl.

Hi all, if I want to simulate a multiphase pipe flow but there is no external rotational flow towards the pipe, should I set to Axisymmetric or Axisymmetric Swirl in tab “General” ?
general tab.PNG

For "Axisymmetric swirl" , how should I determine the "RNG Options" ? How "Swirl Factor" can be set and work for simulation result?
viscous model.PNG

Then under tab "Cell-zone condition", non of the motion or zone being selected, how Fluent will react in default ? Will it be the same as "Axisymmetric" if I set for example under "Reference Frame" ,all Rotational and Translational Velocity are set to zero ?
cell zone condition.PNG

[rezendervp]

Originally Posted by yylaw

Hi all, if I want to simulate a multiphase pipe flow but there is no external rotational flow towards the pipe, should I set to Axisymmetric or Axisymmetric Swirl in tab “General” ?
Attachment 97771

For "Axisymmetric swirl" , how should I determine the "RNG Options" ? How "Swirl Factor" can be set and work for simulation result?
Attachment 97772

Then under tab "Cell-zone condition", non of the motion or zone being selected, how Fluent will react in default ? Will it be the same as "Axisymmetric" if I set for example under "Reference Frame" ,all Rotational and Translational Velocity are set to zero ?
Attachment 97773

You have three four options: 1) A full 3D simulation and let the flow to flow - more computational efforts
And, to reduce the computational efforts, 2D approaches:2) 2D planar (even though a axcisymetric geometry such a pipe - there is a error in the axial direction flow due to the normal area ( sectionnal area) of each Volume of Control. But I see low errors or even none difference - but is flow dependent( simples cases, low Re, etc) . Avoid this because is mathematically wrong.
3) a axissymetric aproach - better tahen 2D Planar because is mathematical and physically consistent. But the angular veloocity component is assumed symmetric ow null - we have only axial and radial components. If this is the case, It is OK to set up this.
4) assisymetric with swirl - this is an "evolution" of the last one, where the angular velocity component is not negligible. This option avoid to consider a full 3D simulation due this physical behavior that must be accounted for.

Now, if you dont know if the swirll is or isn't impornt or present, choose the last, because is more general wheter theta componet is null or not, it doenst matter. If YES ok, If NOT OK also....

Turbulence... well... tubulence is always 3D and unsteady. But 2D is a good aproximation and we work in 2D all time. How good will it be? Who knows?... each case is a case. There is no a general recipe.
RANS is an aproximation of average flow; URANS is the same, with some unsteady behavior; LES is better, but with a high cost; RSM, RANS approach with more equations and physics embeded.

But all can be resumed to a one question: "is the angular velocity component import? " if NOT, axisymmetric, if YES axisymmetric with swirl.




[/INDENT]