CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   Mass accumulation in transient Simulation (https://www.cfd-online.com/Forums/main/79404-mass-accumulation-transient-simulation.html)

Arevik August 23, 2010 07:52

Mass accumulation in transient Simulation
 
I work on a transient simulation in CFX. It is a Taylor-Couette flow in a narrow gap between two concentric cylinders, with the inner cylinder rotating. I get the expected flow phenomena (wavy taylor vortices).
However, the total pressure at the monitoring points constantly rises. The domain imbalance for mass ist ok (0.002%) also the other imbalances. Residuals after 4 iterations per timestep are ok (Max.Residual 10e-4..10-5), only H-Energy doesn't converge (Max. Res.= 10e-2). As boundary conditions at the cylinder ends I set first translational periodicity, I also tried with symmetry or free slip walls. There is no inlet or outlet. It is a laminar flow, however, I also tried with SST turbulence once.
Whatever I tried out, total pressure /mass was rising. Can anyone tell me what is wrong?

CCTech_Pune August 24, 2010 05:35

Hi

What is the time-step you are using. Generally you should choose 5/RPM ass the timestep. And also check if you have residuals in one or two cells only. Have you checked grid dependency?


Thanks
CCTech Pune

Arevik August 24, 2010 06:49

Hi CCTech Pune

Thanks for you ideas.
The grid should be ok, I did a study about this before starting the transient simulation.
My timesteps were much smaller, I used Courant-Nr ~= 1. An increase of timestep by a factor of 5 didn't change anything to my problem. I will set up a new simulation with ~5 dt/rpm and see what happens. Why do you say 5/rpm? Is there a special theory about that?
I think, this will solve the problem. In the meanwhile I finished my simulation of turbulent Taylor vortex flow. there is still some mass accumulation, but total pressure in monitor points didn't rise. There I used much higher timesteps (~1.5/rpm).

Arevik August 26, 2010 09:32

Unfortunately, increasing the timestep didn't help. Total Pressure / mass still rose (despite the closed system). When I wanted to increase it even more, the simulation broke up with errors only few timesteps after the adaption.

ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| c_fpx_handler: Floating point exception: Overflow |

+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Stopped in routine c_fpx_handler: |


Are there any other suggestions, what I could try?

CCTech_Pune August 26, 2010 09:52

Hi

Can you please send me the complete out file to consultancy@cctech.co.in

I could not figure out what could be reason for crash.

Thanks
CCTech Pune


All times are GMT -4. The time now is 00:32.