Fire Modelling in a closed compartment

427 views
Skip to first unread message

Danish

unread,
Jan 14, 2021, 6:00:15 AM1/14/21
to FDS and Smokeview Discussions

Der FDS User, 

I am working on a fire modeling  in a closed hall/ space and based on the FDS user guide for the closed compartment I added the pressure zone in the compartment. After adding the pressure zones I found that the hrr is continuously increasing. 

To test the accuracy of the model I started the simulation where I opened the door leading to the closed hall and provided the open vent at the end of the domain, Case number 1. See the attached image. In that case, I found the HRR as per the defined value in the model.

While for the actually closed compartment scenario, I tested two cases depending upon the location of the pressure zone. Case 2 defines the pressure zone outside the fire origin zone and  Case 3 pressure zone within the fire origin zone. See the attached image. In these two cases hrr continuously increase with time event with fixed HRRPUA value defined in the model.

The mechanical system are as follows

a) Extract = 20 m3/s

b) Supply = 19 m3/s (95%)

Please let us know your thought on how to model such a closed scenario. 

Thanks for your time and suggestions.

Regards

Wali

Capture.PNG

dr_jfloyd

unread,
Jan 14, 2021, 9:01:14 AM1/14/21
to FDS and Smokeview Discussions
If you have a volume with an OPEN vent, then setting a pressure zone in it doesn't do anything as you cannot have a background pressure rise with an OPEN boundary, 

To set a pressure zone in complex volume with the current release you just need to define a single XYZ point somewhere in the zone and FDS will locate all the connected cells.

If your supply and exhaust is in a volume with no OPEN boundaries you need to specify leakage. Without leakage, if you extract 20 m3/s and supply 19 m3/s, then you will start to draw a vacuum and this will eventually lead to problems, If you drop the density below the clipping density due to a vacuum, FDS will clip the density which could be a factor in your HRR. Even if you specified the same volume flow you can run into issues over time if the temperature at the exhaust (and hence the density) is not identical to the supply inlet temperature.

Danish

unread,
Jan 17, 2021, 12:46:16 AM1/17/21
to FDS and Smokeview Discussions

Dear Jason, 

I followed one local_leakage temperature fds example, where the supply vent is applied at one side along with the top and bottom leakage phenomenan are observed with the open boundary at the other end of the domain.

Next, based on you suggestion I updated the model where I defined the pressure zone inside the fire compartment and also remove the open vent from the model but keep the door leakage surface. After performing the simulation still i am getting a same results i.e increase in hrr with time. I considered two cases, see attached images and relevatent fds files.

Please have a look and advice.

 

Regards

Wali

fds_version.PNG
Capture2.PNG
Message has been deleted

Danish

unread,
Jan 17, 2021, 1:56:21 AM1/17/21
to FDS and Smokeview Discussions
Hi Jason, 

Not able to attach the relevant fds files, the file size is around 2Mb. Is there any other way to transfer the files?

Regards

Glenn Forney

unread,
Jan 17, 2021, 3:02:08 AM1/17/21
to fds...@googlegroups.com
Try changing extension to .txt

--
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/c1311173-35fc-48d6-b6d2-253875c9005bn%40googlegroups.com.
Message has been deleted

Danish

unread,
Jan 17, 2021, 7:08:09 AM1/17/21
to FDS and Smokeview Discussions
Hi Jason and Glenn, 

I know it always recommended to keep lesser number obstructions so that FDS experts found the issue quickly, but since my model is complex in shape so I did not make a model in FDS I,  just import the 3d model and convert to blocks therefore, it includes a large number of obnstrcutions. 

Please accept the file and guide how to model the fire scenario with pressure zones and leakage to sustain the defined fire size in the model instead of the growing fire. 

Thanks for your the time to look into the case file.

Regards
Wali
test4.txt
test5.txt

Kevin McGrattan

unread,
Jan 17, 2021, 12:44:07 PM1/17/21
to fds...@googlegroups.com
Use the latest version of FDS. We cannot diagnose problems with this older version.

Danish

unread,
Jan 19, 2021, 5:20:47 AM1/19/21
to FDS and Smokeview Discussions
Dear Kevin,

Thanks, I ran the same two scenarios using the latest FDS version 6.7.5 and now i am getting the stabilized hrr curve (See attached curve). In addition to that I have a few queries can you please guide me:

a)   Like in the current modeling setup, I defined the pressure zone within the fire compartment as a box and defined the leakage surface of 100cm2 at the bottom of the surface without an open vent in the domain, ideally to define the boundary conditions within a CFD domain we need at least one pressure boundary conditions since we don't have an open vent in the domain how the software use this pressure bc.

b)  I am not getting any flow at the door leakage,  See attached image 4, Is this because I put volume flow device not volume flow wall device. 

c) Since in my domain there is no open vent, then how FDS works to remove or add extra air into the domain due to differences in supply and extract flow rate.

Thanks for the time to look into the matter.
Regards
Wali

Capture3.PNG
Capture4.PNG

Kevin McGrattan

unread,
Jan 19, 2021, 10:27:35 AM1/19/21
to fds...@googlegroups.com
a)   Like in the current modeling setup, I defined the pressure zone within the fire compartment as a box and defined the leakage surface of 100cm2 at the bottom of the surface without an open vent in the domain, ideally to define the boundary conditions within a CFD domain we need at least one pressure boundary conditions since we don't have an open vent in the domain how the software use this pressure bc.

Even if there are no OPEN boundaries, FDS still implicitly defines "ZONE 0" as the infinitely large exterior domain. 

b)  I am not getting any flow at the door leakage,  See attached image 4, Is this because I put volume flow device not volume flow wall device. 

I do not know. Try it. Better yet, create a very simple case with a sealed room and leaking door. Ensure that your prescription of leakage is correct by monitoring pressure and flow.

c) Since in my domain there is no open vent, then how FDS works to remove or add extra air into the domain due to differences in supply and extract flow rate.

See my first answer.
Reply all
Reply to author
Forward
0 new messages