Strange Density Contours when undertaking both steady and transient calcs

mgregos · March 27, 2023, 12:00

Hi,

I am having issues getting density to converge in both steady state and transient problems, as seen in the image below.

My mesh is pretty good, mesh skewness 0.8, orthogonal quality >0.2 and aspect ratio max of circa 13.

could anybody advise on what would cause this please? If you need more infor please let me know.

Thanks!

NickFL · March 27, 2023, 14:16

What exactly are you expecting to see here. I see a 1% change between various points in the flow field, so basically just a little numerical noise. Is there a physical phenomenon that should be present? Without more knowledge of the problem, we cannot recommend anything.

mgregos · March 28, 2023, 04:46

Hi Nick,

Thanks for your reply. I would expect to see steady density in this region but a change in density further along the fluid flow. I've attached a re-run seen below for steady state, the initial conditions are a mass flow of 1kg/s at inlet mass flow and 1kg/s at the outlet for mass flow. I would expect to see a change in density in front of the obstruction but in this case the density is staying constant throughout. If you need anything else please let me know.1kgs desnity.PNG

1kgs pressure.PNG

1kgs vel.PNG

Mass Flow Rate.PNG

Residuals 1kgs.PNG

NickFL · March 28, 2023, 05:18

You are using a mass-flow at both the inlet and outlet? Generally we avoid that.

Why are you expecting a measurable density change with such a small velocity? Based on what little we know, if this is at room temperature air I would estimate this as Ma < 0.1 which is clearly incompressible and a constant density air would be sufficient.

As for the residuals and the flow-rate of the x-velocity, I would say neither of them are really converged. The latter is showing a ~15% change from the peak and accelerating. The contours of the x-velocity show that in the near-wall region there looks to be some grid-size effects.

But to your original question of the density contours, basically I don't see a reason to move away from constant density which makes your problem go away.

mgregos · March 28, 2023, 09:36

Hi Nick,

Yes I added it maybe thinking that only having an outflow was affecting the density, however it made a negligible difference. I will remove it again now.

Thank you yes that is very clear, it is just the strange shaping of the density changes that made me seek further clarification.

Yes agreed, thank you very much for your help.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Steady or transient which is faster to get thermal steady status	hogglife	OpenFOAM Running, Solving & CFD	1	January 3, 2023 03:24
Transient simulation gives worse results than steady state simulation	jgross	CFX	12	January 21, 2021 12:21
The difference between steady state and transient	JuPa	CFX	36	December 9, 2019 22:50
2way FSi, Initialize with steady solution, Fluent, Transient Sturcural, System Coupli	mmkkeshavarzi	FLUENT	0	June 22, 2016 08:26
Steady/ Transient Calculations	Saidul	ANSYS	0	May 4, 2015 19:15

March 27, 2023, 14:16		#2
NickFL Senior Member Join Date: Jun 2009 Location: Technische Universität Chemnitz Posts: 107 Rep Power: 16	What exactly are you expecting to see here. I see a 1% change between various points in the flow field, so basically just a little numerical noise. Is there a physical phenomenon that should be present? Without more knowledge of the problem, we cannot recommend anything.

March 28, 2023, 04:46		#3
mgregos New Member Greg Join Date: Mar 2023 Posts: 3 Rep Power: 3	Hi Nick, Thanks for your reply. I would expect to see steady density in this region but a change in density further along the fluid flow. I've attached a re-run seen below for steady state, the initial conditions are a mass flow of 1kg/s at inlet mass flow and 1kg/s at the outlet for mass flow. I would expect to see a change in density in front of the obstruction but in this case the density is staying constant throughout. If you need anything else please let me know.1kgs desnity.PNG 1kgs pressure.PNG 1kgs vel.PNG Mass Flow Rate.PNG Residuals 1kgs.PNG

March 28, 2023, 05:18		#4
NickFL Senior Member Join Date: Jun 2009 Location: Technische Universität Chemnitz Posts: 107 Rep Power: 16	You are using a mass-flow at both the inlet and outlet? Generally we avoid that. Why are you expecting a measurable density change with such a small velocity? Based on what little we know, if this is at room temperature air I would estimate this as Ma < 0.1 which is clearly incompressible and a constant density air would be sufficient. As for the residuals and the flow-rate of the x-velocity, I would say neither of them are really converged. The latter is showing a ~15% change from the peak and accelerating. The contours of the x-velocity show that in the near-wall region there looks to be some grid-size effects. But to your original question of the density contours, basically I don't see a reason to move away from constant density which makes your problem go away.

March 28, 2023, 09:36		#5
mgregos New Member Greg Join Date: Mar 2023 Posts: 3 Rep Power: 3	Hi Nick, Yes I added it maybe thinking that only having an outflow was affecting the density, however it made a negligible difference. I will remove it again now. Thank you yes that is very clear, it is just the strange shaping of the density changes that made me seek further clarification. Yes agreed, thank you very much for your help.