[killsecond]

Originally Posted by TianC

Hi guys,

To my understanding, the wallShearStress utility calculates the wall shear stress at the latest timestep. It is a post-processing utility.

wallShearStress can be a good way to monitor if a simulation has settled down to a steady state. Ideally you don't want to be saving all your data for many timesteps until the simulation has settled (since this consumes vast amounts of data as you may have noticed!)

So my question is:

Is there any way of outputting wallShearStress at every timestep (i.e. not having to use the utility for just the last timestep every time)?

Ideally, is it possible of outputting JUST the wallShearStress at more frequent intervals than you output your other data?

Also, is it possible to do some kind of average magnitude to see how the wall shear stress behaves?

Tian Coulsting，

Have you solved these problems now? I need to do these now, help me. Thank you!

[Shahr]

You can use utilityFunctionObjects for this purpose. Google/grep it for usage guide. There you can adjust the output frequency as well.

I want to add that although judging by the magnitude of WSS might be a good way of judging about the statistical convergence of the simulation, I have found it only a necessary conditions, not sufficient. You have to compare probes' values of higher order statistics, time-averaged profiles, etc.

[killsecond]

Originally Posted by Shahr

You can use utilityFunctionObjects for this purpose. Google/grep it for usage guide. There you can adjust the output frequency as well.

I want to add that although judging by the magnitude of WSS might be a good way of judging about the statistical convergence of the simulation, I have found it only a necessary conditions, not sufficient. You have to compare probes' values of higher order statistics, time-averaged profiles, etc.

Hi Shahr,

Thank you very much for your advice, I will try. In fact, I want to show the skin friction coefficient distribution around the cylinder, however the flow around the cylinder is unsteady, so I need to show the time-averaged profiles.

[wyldckat]

Originally Posted by huangxianbei

Hi,Bruno
My case is a 90 bend duct with an aspect ratio of 6. Smagorinsky model is used with mean y+ <4, it's incompressible.
The inlet condition is obtain by interpolate the experimental velocity, and then mapped to the grid points.
The mesh's max aspect ratio is <700, which is OK when doing the checkMesh
Here is the comparison of the two results at inletAttachment 34454
The first 7 rows of data are obtained using sample line, the 1st is distance, and following are the wallGradU vector and wallShearStress vector, each has 3 components. The x direction is the studied direction
Xianbei

Hi Xianbei,

I've had this post of yours on my to-do list and only now was I able to look into this.
I have to say that I was not able to make heads or tails from the data you've provided Reproducibility is critical for when we need to diagnose problems.

Anyway, from what I've seen in later posts of yours, it seems that you've solved the problem, as indicated here: http://www.cfd-online.com/Forums/ope...tml#post516964 post #21

Which means that there isn't much I can do here... oh well...

Best regards,
Bruno

[gu1]

Originally Posted by killsecond

Hi Shahr,

Thank you very much for your advice, I will try. In fact, I want to show the skin friction coefficient distribution around the cylinder, however the flow around the cylinder is unsteady, so I need to show the time-averaged profiles.

Have you solved these problem? I need to do these now.
Thanks