Numerical Instability in FDS 5

[Nat Ong]

Hi,

I am currently running a fds model to model the impact of wind in a building which is on fire. The wind speed is of 3.24m/s and extra domain is taken into account to model its  impact to the surrounding as well. However, my model kept facing numerical instability error message which I haven't seen it in FDS5. It's common to tackle with this error message in FDS4 however it surprised me in FDS5.

I have my fire delayed for 5 min just to consider the turbulence caused by wind inside the building and I don't think this is the reason my model stopped. I have tried change grids several time ie:  to grid the wind panel in 1 mesh, which believed to cause this numerical instable error AND regrid the model to have a ratio of 1:2 throughout all the axis. However, all models failed with numerical instability error again.

Please can anyone advise if they have encounter this issue before?

Attached is the input file to the fds model.

Thanks,

Natalie

[Nat Ong]

[Dave McGill]

Hi Nat,

When in the simulation does it fail? Do you see any areas of very high velocity at the point? I suspect the problem is the flat meshes that are stacked on top of each other. Without knowing when it fails here's how I would trouble-shoot it.

1. Get rid of the building and the fire and just run it with the wind.
   
2. Add the building back in.
3. Add the fire back in.

I also suggest you use the MEAN_FORCING approach to create the wind. See, https://groups.google.com/forum/?fromgroups=#!searchin/fds-smv/wind/fds-smv/tYGSl_32BAo/31_PhyaENfAJ

Dave

[dr_jfloyd]

I would recommend you get rid of the meshes with only a handful of grid cells in the z-direction by merging them into another mesh (for example meshes 19 and 20 have the same resolution but mesh 20 is only 4 grid cells high).   I would also recommend  that you attempt merge some of the other meshes as well.  For example mesh 27 and 28 have the same xy footprint and the same resolution , joining them together creates a mesh that is still less than half your largest mesh and if you are running in parallel it is those largest meshes that will dictate the run time. 

[Mohamed ASSAL]

Hi Natalie,

How did you specify the wind velocity i checked your input i can't see
any wind source..3.24m/s.

Cheers Mohamed.

2013/2/22, dr_jfloyd <drjf...@gmail.com>:

> I would recommend you get rid of the meshes with only a handful of grid
> cells in the z-direction by merging them into another mesh (for example
> meshes 19 and 20 have the same resolution but mesh 20 is only 4 grid cells
> high). I would also recommend that you attempt merge some of the other
> meshes as well. For example mesh 27 and 28 have the same xy footprint and
> the same resolution , joining them together creates a mesh that is still
> less than half your largest mesh and if you are running in parallel it is
> those largest meshes that will dictate the run time.
>
>
> On Friday, February 22, 2013 7:08:50 AM UTC-5, Dave McGill wrote:
>>
>> Hi Nat,
>>
>> When in the simulation does it fail? Do you see any areas of very high
>> velocity at the point? I suspect the problem is the flat meshes that are
>> stacked on top of each other. Without knowing when it fails here's how I
>> would trouble-shoot it.
>>

>> 1. Get rid of the building and the fire and just run it with the wind.
>>
>> 2. Add the building back in.
>> 3. Add the fire back in.

>>
>> I also suggest you use the MEAN_FORCING approach to create the wind. See,
>>
>> https://groups.google.com/forum/?fromgroups=#!searchin/fds-smv/wind/fds-smv/tYGSl_32BAo/31_PhyaENfAJ
>>
>> Dave
>>
>

> --
> 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/msg/fds-smv/-/skEsleQYrg8J.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>