wind simulation: unsual wind velocity pattern in fine mesh area

[FDS_user]

Hi,

I am using M-O wind profile with the following values:

&WIND SPEED=5.0, Z_REF=5.0, DIRECTION=270.0, L=1000000.0 , Z_0=0.0002 /
&VENT PBX= -140.8, SURF_ID='OPEN' /
&VENT PBX= 396.8, SURF_ID='OPEN' /
&VENT PBY= -140.8, SURF_ID='PERIODIC FLOW ONLY', COLOR='IVORY'/
&VENT PBY= 140.8, SURF_ID='PERIODIC FLOW ONLY', COLOR='IVORY' /
&VENT PBZ= 0.0, SURF_ID='GROUND' /
&VENT PBZ=153.6, SURF_ID='OPEN' /

&SURF ID= 'GROUND'/

In the place where I have fine meshes (0.4m), I get unusual wind velocity patterns.

Can anyone please guide how to get away with such unusual wind velocity patterns?

[Randy McDermott]

We need to see the input file, or at least to see the grid resolution in that area.  Looks like you have some ringing at a coarse-fine mesh interface?

--
You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/5f1ebb56-2f5f-44b5-b11d-9d8ab9805500n%40googlegroups.com.

[FDS_user]

Thanks for the response; the input file is attached to this message!

Can you also please let me know if such a mesh configuration is advisable? I have fire meshes of 0.4m, and around it 0.8m, then 1.6m and finally away from fire mesh is 3.2m! All meshes are almost symmetrical around the x axis and the  y axis.

[FDS_user]

The unusual pattern develops when the wind enters from 0.8m mesh to 0.4m!

[Kevin McGrattan]

I am running the case.

[Kevin McGrattan]

I don't think there is a problem here. The "wiggles" in the velocity field you see on the finer meshes are the result of better resolution and less numerical viscosity that tends to dampen physical oscillations in the flow. If you are interested in developing a realistic atmospheric boundary layer, then you should use a consistent mesh throughout so as not to dampen the naturally-occuring turbulence. 

[FDS_user]

Thanks for the reply! As per the suggestion, I may have to go for consistent meshes.

[FDS_user]

Hello Kevin,

I changed the meshes and made consistent mesh right from the entry of the wind from the left side of the domain. The place/meshes where ringings are happening are of size 0.4m, and above it is a mesh of 0.8m. If you see the image attached, the red mesh is of 0.4m, the blue is of 0.8m, the green mesh is of 1.6m, and the yellow mesh is 3.2m. The red mesh, roughly after 100m from the left side of the domain, will have the fire. I have also attached the file. Can you please look into it and give some suggestions? Is it that I need to have consistent meshes of 0.4m all through the vertical direction also?

Thanks!

[FDS_user]

Addition: I made consistent meshes from the left side of the domain for a fairly long distance of roughly 200m!

[Randy McDermott]

There is going to be ringing in the pressure solution at these resolution jumps.  If you cannot tolerate this error, then you need to stretch the bottom mesh such that the resolutions match at the mesh boundary.  Usually, this allows for fewer cells anyway.

--

You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/823e0393-3b0c-4369-848b-a3f7f0c0cdfbn%40googlegroups.com.

[FDS_user]

Thanks, I will try to implement the stretching thing to the bottom mesh!