[bany]

When i use the mapped BC to obtain a fully developed circular pipe flow in LES with pisoFoam, i get a false u profile along the radial direction which increases with time.
And in this case, my y+~5, and the offset is set to (0.025 0 0).The length of the pipe is 30mm and the initial velocity is (110 0 0). I set the setAverage to true to maintain the pipe flow. To ensure the Re being 10500, i set the nu=1.57e-5.
Can anyone tell me why i cannot obtain a fully developed circular pipe flow with those setup.
Any advice are appreciated!

[guanjiang.chen]

Originally Posted by bany

When i use the mapped BC to obtain a fully developed circular pipe flow in LES with pisoFoam, i get a false u profile along the radial direction which increases with time.
And in this case, my y+~5, and the offset is set to (0.025 0 0).The length of the pipe is 30mm and the initial velocity is (110 0 0). I set the setAverage to true to maintain the pipe flow. To ensure the Re being 10500, i set the nu=1.57e-5.
Can anyone tell me why i cannot obtain a fully developed circular pipe flow with those setup.
Any advice are appreciated!

Did you use fixedValue boundary condition for the initial vilocity?

[bany]

Originally Posted by guanjiang.chen

Did you use fixedValue boundary condition for the initial vilocity?

Oh, thanks a lot! Surely, i had try a fixedValue BC, however, i cannot obtain a turbulent pipe flow(the U profiles looks like a laminar flow) expect for one situation i.e. y+<1. But the case for y+<1, the number of mesh is so many.
In addition, i try to use the wallfunctions. Unfortunately, i also cannot get a fully turbulent pipe flow.
For more details, you can look up the attachments! All cases can be run by blockMesh ——> pisoFoam

[guanjiang.chen]

Originally Posted by bany

Oh, thanks a lot! Surely, i had try a fixedValue BC, however, i cannot obtain a turbulent pipe flow(the U profiles looks like a laminar flow) expect for one situation i.e. y+<1. But the case for y+<1, the number of mesh is so many.
In addition, i try to use the wallfunctions. Unfortunately, i also cannot get a fully turbulent pipe flow.
For more details, you can look up the attachments! All cases can be run by blockMesh ——> pisoFoam

Hi bany,

If I set, I will also set boundary layer like you. I don't know whether calculating for longer time can show turbulent pipe flow.
Did do some Urans case and in Urans case is there turbulent flow?
What I think can try maybe using cyclic boundary for inlet and outlet.
Making nearwall mesh better is also a choice.

Sincerely,
Guanjiang

[bany]

Originally Posted by guanjiang.chen

Hi bany,

If I set, I will also set boundary layer like you. I don't know whether calculating for longer time can show turbulent pipe flow.
Did do some Urans case and in Urans case is there turbulent flow?
What I think can try maybe using cyclic boundary for inlet and outlet.
Making nearwall mesh better is also a choice.

Guanjiang, thanks for your applying. I think that the calculating time is not the main reason.Because, when i run these cases, about 20 flow times are guaranteed.
In addition, i have tried to use the cyclic BC with the gradp to force the flow. And by this method, we can get a fully turbulence.
However, i do not know why i cannot get a fully turbulence by mapped BC or wallfuncions in LES.
Whatever, thanks for your attention.

[guanjiang.chen]

Originally Posted by bany

Guanjiang, thanks for your applying. I think that the calculating time is not the main reason.Because, when i run these cases, about 20 flow times are guaranteed.
In addition, i have tried to use the cyclic BC with the gradp to force the flow. And by this method, we can get a fully turbulence.
However, i do not know why i cannot get a fully turbulence by mapped BC or wallfuncions in LES.
Whatever, thanks for your attention.

Did you compare pitzDailyMapped and pitzDaily? /newhome/va19337/OpenFOAM/OpenFOAM6/tutorials/incompressible/pisoFoam/LES/pitzDailyMapped

They shows different results. The mapped boundary condition is special I think.

[bany]

Originally Posted by guanjiang.chen

Did you compare pitzDailyMapped and pitzDaily? /newhome/va19337/OpenFOAM/OpenFOAM6/tutorials/incompressible/pisoFoam/LES/pitzDailyMapped

They shows different results. The mapped boundary condition is special I think.

OK, i noticed both cases. However, you can notice that in the extended square 2-D pipe, the turbulence seems to be non-existent. That is, the turbulence exists in the container rearwards and in my opinion, it is formed due to the shear between the slow flow in container rearwards and the fast flow from the square 2-D pipe.
Thanks for your continuous answer!

[guanjiang.chen]

Originally Posted by bany

OK, i noticed both cases. However, you can notice that in the extended square 2-D pipe, the turbulence seems to be non-existent. That is, the turbulence exists in the container rearwards and in my opinion, it is formed due to the shear between the slow flow in container rearwards and the fast flow from the square 2-D pipe.
Thanks for your continuous answer!

There is much small turbulence in pitzDailyMapped I think. Let's check more material about Mapped boundary before use it.

[bany]

Originally Posted by guanjiang.chen

There is much small turbulence in pitzDailyMapped I think. Let's check more material about Mapped boundary before use it.

Okay, thanks a lot!

[H.Tofighian]

Originally Posted by bany

When i use the mapped BC to obtain a fully developed circular pipe flow in LES with pisoFoam, i get a false u profile along the radial direction which increases with time.
And in this case, my y+~5, and the offset is set to (0.025 0 0).The length of the pipe is 30mm and the initial velocity is (110 0 0). I set the setAverage to true to maintain the pipe flow. To ensure the Re being 10500, i set the nu=1.57e-5.
Can anyone tell me why i cannot obtain a fully developed circular pipe flow with those setup.
Any advice are appreciated!

Hi.
What is your divScheme for velocity? Maybe the numerical diffusion damps eddies?
Also, eddies formation stars from the boundary layer, so you need to resolve it.

[bany]

Originally Posted by H.Tofighian

Hi.
What is your divScheme for velocity? Maybe the numerical diffusion damps eddies?
Also, eddies formation stars from the boundary layer, so you need to resolve it.

You can find more details on the #3 from the attachments.
You can see in the pipe-wallfunction where i set wall functions for nut and k in order to resolve the eddies near the wall. However, the wall functions seem to be invalid. By the way, the wall functions, i think, is fitted with the y+.

In short, thanks a lot.

[lizzy]

Hi bany,

I just encounter the same problem with you: the "laminar" mean velocity profile after using mapped BC at the inlet. Have you found out the solution?

BTW, besides mapping U, do you think it is necessary to map other parameters, such as nut and k...?

Appreciate anyone who can give some advice on this issue. Thanks!

Originally Posted by bany

When i use the mapped BC to obtain a fully developed circular pipe flow in LES with pisoFoam, i get a false u profile along the radial direction which increases with time.
Any advice are appreciated!

[sinatahmooresi]

Originally Posted by bany

When i use the mapped BC to obtain a fully developed circular pipe flow in LES with pisoFoam, i get a false u profile along the radial direction which increases with time.
And in this case, my y+~5, and the offset is set to (0.025 0 0).The length of the pipe is 30mm and the initial velocity is (110 0 0). I set the setAverage to true to maintain the pipe flow. To ensure the Re being 10500, i set the nu=1.57e-5.
Can anyone tell me why i cannot obtain a fully developed circular pipe flow with those setup.
Any advice are appreciated!

Did you find out the reason? I am applying mapped B.C. on pipe flow for a LES simulation and the results inside the pipe does not look instantaneous to me.

[JM27]

Originally Posted by sinatahmooresi

Did you find out the reason? I am applying mapped B.C. on pipe flow for a LES simulation and the results inside the pipe does not look instantaneous to me.

Hi,

I am also working on a similar problem, maybe we can help each other. Do you apply the mapped BC to the U field only or also to nut and k?

Thanks,
J

[guanjiang.chen]

Originally Posted by JM27

Hi,

I am also working on a similar problem, maybe we can help each other. Do you apply the mapped BC to the U field only or also to nut and k?

Thanks,
J

Hi J,

I have abandoned the mapped BC. What I do is conducting the RANS simulation first and then set the RANS results as the initial set for the LES.

Regards,
Guanjiang

[JM27]

Hi Guanjiang,

Thanks for the prompt response. What boundary conditions do you then use in your LES if not using the mapped BC?

[guanjiang.chen]

Originally Posted by JM27

Hi Guanjiang,

Thanks for the prompt response. What boundary conditions do you then use in your LES if not using the mapped BC?

Hi,

What I do is a basic problem of incompressible flow past a cylinder. I fixed the velocity at the inlet and fixed the pressure at the outlet. For the nut, I just use zeroGradient at the inlet and outlet. There are many boundary matches tutorial online. You can try and get a best one.

Regards,
Guanjiang

[bany]

Originally Posted by JM27

Hi Guanjiang,

Thanks for the prompt response. What boundary conditions do you then use in your LES if not using the mapped BC?

The fully developed turbulence in LES can be obtained by refinemesh in the wall regions. It is the only way where i got a fully developed turbulence with LES. However, the refined mesh near the wall will lead to more time cost.

Any advice will be appreciated!

[Junrui]

Originally Posted by bany

The fully developed turbulence in LES can be obtained by refinemesh in the wall regions. It is the only way where i got a fully developed turbulence with LES. However, the refined mesh near the wall will lead to more time cost.

Any advice will be appreciated!

Hi bany,
I am now also working on LES for fully developed turbulent channel flow. However, even my y+ is controlled below 2, the velocity profile indicates that the flow might not be fully developed turbulent flow. If convenient, could you please tell me about your boundary conditions, the difference scheme for the convective term, and how many circulations have you run to get the fully developed turbulent flow?
Thanks for your time,
Regards,
Junrui

[sinatahmooresi]

Originally Posted by Junrui

Hi bany,
I am now also working on LES for fully developed turbulent channel flow. However, even my y+ is controlled below 2, the velocity profile indicates that the flow might not be fully developed turbulent flow. If convenient, could you please tell me about your boundary conditions, the difference scheme for the convective term, and how many circulations have you run to get the fully developed turbulent flow?
Thanks for your time,
Regards,
Junrui

Hello Junrui,

In addition to the mesh refinement adjacent to the wall, I suggest you to switch to cubic convective schemes in your fvSchemes dictionary. Using this, I could solve the problem of turbulence generation in my pipe flow case.



Sina