Mesh boundaries near critical actions

[Erik Carlsson]

Hi all,

I have always built my FDS models with no mesh boudnaries near "critical actions" such as fires, high velocity fields etc.  This is still stated as a mesh alignment rule of thumb in the user's guide:

"Avoid putting mesh boundaries where critical action is expected, especially fire. Sometimes fire spread from mesh to mesh cannot be avoided, but if at all possible try to keep mesh interfaces relatively free of complicated phenomena since the exchange of information across mesh boundaries is not yet as accurate as cell to cell exchanges within one mesh."

I was recently told however that this is not as much an issue anymore as it used to be.  Can anyone shed any light on this for me?  The gut feel is to still avoid mesh boundaries in areas with high velocities, combustion etc.

Thanks,

Erik

[dr_jfloyd]

Mesh-to-mesh exchanges are significantly improved in the more recent releases as compared to FDS 5. It is still true that the exchanges are not as accurate as being fully within one mesh. You do now have some control over that accuracy using the VELOCITY_TOLERANCE, MAX_PRESSURE_ITERATIONS, and RELAXATION_FACTOR inputs on PRES. My experience is that for simple pool fire type simulations, putting a mesh boundary through the pool fire isn't the issue that it used to be. It is still a good idea to avoid it, but not critical.

[Kevin]

We still see alot of non-optimal meshing in submitted bug reports. A popular one is to create 4 to 8 cell high horizontal layers stacked on top of each other within the same compartment. Sure, this divides up the work, but it creates the maximum area of interfaced mesh boundaries. So we continue to caution users because we just can't anticipate what will be done.

[Randy McDermott]

I use mesh boundaries within a fire routinely.  The bigger issue is using a refinement ratio near the action.  Look at the Sandia_Plumes and UMD_Line_Burner examples in the validation suite.  Also, in Sec 6.4.10 of the user guide we note that it is a good idea to use SECOND_ORDER_INTERPOLATED_BOUNDARY=.TRUE. on MISC for multi-mesh calculations when the mesh boundary is away from an obstruction.  We are working on making this default.

--
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 post to this group, send email to fds...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/212c75ee-2d74-4048-9fd8-da159a8bc4c7%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.