Different Pressure in steady state and tranisente simulation

abcdefgh · January 28, 2015, 10:16

Hello,

I hope you can help me. I currently simulate a pressure drop within a tube with an obstacle in it. When I calculate the steady state model there is a completely different pressure drop out (much higher) than if I have a tranisent simulation. The lower value would agree much better with reality (messure). How can the high value arrive at steady state?
The transient calculation I have started from the starting points of steady state results.
Thanks in advance and best regards.

Opaque · January 28, 2015, 11:50

Are both solution similarly converged ?

That is, the residuals are below a meaningful threshold, and the transient solution is no longer changing with residuals at the same level as the steady solution. If the residual levels are different (within reason), one of them has not converged.

ghorrocks · January 28, 2015, 16:32

Also have you run the transient simulation long enough so that it converges to a steady state solution? It might be still int he startup transients.

abcdefgh · January 29, 2015, 01:40

Thanks for the replies.
The difference can really be because of the convergence. The tranisent calculation gets not under the value 1.0e-05 and stays at a constant level (RMS P-measure, RMS U-Mom, RMS Mom-V , RMS W-Mom). In the steady state calculation the Varibable Value of Momentum and Mass was achieved a level under 1.0e06 in less than 100 steps.
Does that mean I can believe the result of the steady state calculation more? What can I do to make the tranisent calculation converges as well? And how can it generally be that the pressure loss of the calculation (steady state) is much too high in comparison to reality?

Thanks so much !!!

ghorrocks · January 29, 2015, 16:47

Have you done all the normal accuracy checks? See FAQ: http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

The mesh sensitivity is a very important one. If you simulation is inaccurate it will just be producing random numbers and comparing one simulation to another when they are random numbers is pointless. You have to ensure your simulation is accurate before it is worth comparing it with anything.

January 28, 2015, 10:16	Different Pressure in steady state and tranisente simulation	#1
abcdefgh New Member Join Date: Oct 2014 Posts: 24 Rep Power: 11	Hello, I hope you can help me. I currently simulate a pressure drop within a tube with an obstacle in it. When I calculate the steady state model there is a completely different pressure drop out (much higher) than if I have a tranisent simulation. The lower value would agree much better with reality (messure). How can the high value arrive at steady state? The transient calculation I have started from the starting points of steady state results. Thanks in advance and best regards.

January 28, 2015, 11:50		#2
Opaque Senior Member Join Date: Jun 2009 Posts: 1,804 Rep Power: 32	Are both solution similarly converged ? That is, the residuals are below a meaningful threshold, and the transient solution is no longer changing with residuals at the same level as the steady solution. If the residual levels are different (within reason), one of them has not converged.

January 28, 2015, 16:32		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Also have you run the transient simulation long enough so that it converges to a steady state solution? It might be still int he startup transients.

January 29, 2015, 01:40		#4
abcdefgh New Member Join Date: Oct 2014 Posts: 24 Rep Power: 11	Thanks for the replies. The difference can really be because of the convergence. The tranisent calculation gets not under the value 1.0e-05 and stays at a constant level (RMS P-measure, RMS U-Mom, RMS Mom-V , RMS W-Mom). In the steady state calculation the Varibable Value of Momentum and Mass was achieved a level under 1.0e06 in less than 100 steps. Does that mean I can believe the result of the steady state calculation more? What can I do to make the tranisent calculation converges as well? And how can it generally be that the pressure loss of the calculation (steady state) is much too high in comparison to reality? Thanks so much !!!

January 29, 2015, 16:47		#5
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Have you done all the normal accuracy checks? See FAQ: http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F The mesh sensitivity is a very important one. If you simulation is inaccurate it will just be producing random numbers and comparing one simulation to another when they are random numbers is pointless. You have to ensure your simulation is accurate before it is worth comparing it with anything.