CFD Online Discussion Forums
Page 2 of 2 1 2
Show 40 post(s) from this thread on one page

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Strange high velocity in centrifugal pump simulation (https://www.cfd-online.com/Forums/openfoam-solving/138495-strange-high-velocity-centrifugal-pump-simulation.html)

huangxianbei July 15, 2014 03:39
Quote:
Originally Posted by tomf (Post 501509)
Hi,

I have not performed any real LES cases, just played around a bit, so I can not say I have seen something similar.

Tom
Hi,Tom:
Thank you for your reply, I think this is quite unusual, the yPlus should not vary so much between RANS and LES. I'll go on digging the reason:(:(

Xianbei

kiddmax August 12, 2014 12:37
Hi Xianbei,

I am wondering how do you set your initial condition for SRFSimpleFoam. Since I always get the computation divergent after a few steps. That the continuity errors become quite large. Do you have any idea how to overcome this problem?

Thanks a lot!

Best regards
Ye

huangxianbei August 12, 2014 20:35
Quote:
Originally Posted by kiddmax (Post 505616)
Hi Xianbei,

I am wondering how do you set your initial condition for SRFSimpleFoam. Since I always get the computation divergent after a few steps. That the continuity errors become quite large. Do you have any idea how to overcome this problem?

Thanks a lot!

Best regards
Ye
Hi, Ye:
All the conditions are in the former replies. You should use 10-20 steps in one PIMPLE loop, which is
nOuterCorrectors
Hope it helps

Xianbei

kiddmax August 14, 2014 03:20
Hi Xianbei,

Thank you so much for your quick reply. I see your conditions setting. I will try my case further.
I still have some questions:

1. Did you use only one channel or full impeller for your successful simulation?

2. Is you simulation RANS or LES? How can the mesh is so less? only 0.3 million?

3. Have you ever tried MRFSimpleFoam? Do you think only one channel or full impeller works by using MRFSimpleFoam solver?

Thank a lot again.

Best regards
Ye

huangxianbei August 15, 2014 01:39
Quote:
Originally Posted by kiddmax (Post 505898)
Hi Xianbei,

Thank you so much for your quick reply. I see your conditions setting. I will try my case further.
I still have some questions:

1. Did you use only one channel or full impeller for your successful simulation?

2. Is you simulation RANS or LES? How can the mesh is so less? only 0.3 million?

3. Have you ever tried MRFSimpleFoam? Do you think only one channel or full impeller works by using MRFSimpleFoam solver?

Thank a lot again.


Best regards
Ye
1/ one channel
2/ LES, structured mesh is used
3/ MRFSimpleFoam, this can be used in both one channel and full impeller

RodriguezFatz August 15, 2014 02:14
Quote:
Originally Posted by huangxianbei (Post 500668)
I want to do LES calculation with the same Mesh. In fact, after computing with RANS, the y+ turns out to be <2.5 around the blade, so it's fine enough to do LES.
Hi Xianbei, from the fact y+ < 2.5 you can not conclude that the mesh is capable of being used for LES. LES also has strong requirements on the boundary mesh in z+ and x+ directions. That is the main reason why LES is computationally so expensive.

huangxianbei August 15, 2014 02:27
Quote:
Originally Posted by RodriguezFatz (Post 506086)
Hi Xianbei, from the fact y+ < 2.5 you can not conclude that the mesh is capable of being used for LES. LES also has strong requirements on the boundary mesh in z+ and x+ directions. That is the main reason why LES is computationally so expensive.
Hi,Philipp:
Thank you for your reminding. In fact, the mesh now is further improved to be capable for LES. The mesh in the two directions is refined with totally 1 million grids. However, the result is not good comparing with the result obtained in Fluent with the same mesh. ,especially near the pressure side. I changed the pisoFoam by add Coriolis force and centrifugal force instead of using MRFSimpleFoam to speed up the calculation. I think it's the near wall treatment that cause the deviation. However, the only wallfuntion( used on hub and shroud whose y+>30) is nuSgsSpaldingWallFunction , no further near wall models can be chosen:(

Xianbei


All times are GMT -4. The time now is 20:43.
Page 2 of 2 1 2
Show 40 post(s) from this thread on one page