Radiative Heat Flux on a Burning Surface in FDS

446 views
Skip to first unread message

Priscilla Mariani

unread,
Feb 12, 2018, 9:15:14 PM2/12/18
to FDS and Smokeview Discussions
Hi,

I intend to simulate a burning timber facade element with a certain net heat flux set on its burning surface to see the effect on its adjacent timber facade elements, whether it radiates enough to cause piloted ignition on the adjacent facade when set at a certain distance.
However, we can't seem to get the expected level of radiative heat flux produced by the burning surface. 
The HRRPUA is set at 110kW/m2 and the radiactive fraction is re-stated as 35%, so we are expecting to see ~38.5kW/m2 radiative heat flux on the burning surface, but the boundary output on it only shows around 15-25kW/m2 radiative heat flux on average with some little spots of 30kW/m2. The incident heat flux on the adjacent facade has thus become too small from what I expected at the set distance.
Also, we expected the radiative heat flux to increase as we go up the facade element, but it gives out a patchy pattern instead of the expected gradient.

I'm trying to understand what is going on and how I can aptly simulate the 35% radiative heat flux.
Could anyone please help?
The input file is attached.

On another note, if anyone has any suggestion on the radiative fraction and soot yield for a pine-type wood, please let me know. 
I couldn't find these from SFPE so I'm just using red oak soot yield and the FDS default radiative fraction, not sure if it's representative enough.
Thank you! :)
fac_01e.fds

Randy McDermott

unread,
Feb 13, 2018, 8:21:03 AM2/13/18
to FDS and Smokeview Discussions
There is a well-known issue with the discrete angle finite volume radiation solver sometimes called the "star pattern".  It is due to low angular resolution in the radiation solver.  Read up on the parameters on the RADI line.  I would use something like 400 solid angles (default is 105).  But you may need more.  You can adjust the time step parameters so as to keep the workload the same as the 105 angle case.  An example we use in the FDS validation guide to get very good agreement with heat flux away from the flame is given here:



--
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+unsubscribe@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/18dc6d7f-79b3-4d54-8ff3-2b02d3c32d8a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Priscilla Mariani

unread,
Feb 18, 2018, 9:14:06 PM2/18/18
to FDS and Smokeview Discussions
Hi Randy,

I've re-run the model with adjusted radi angle and time step as in the attached file.
However, the results still did not agree with the expectation (result was still lower).
Have I adjusted it correctly?

Thanks.

On Tuesday, February 13, 2018 at 9:21:03 PM UTC+8, Randy McDermott wrote:
There is a well-known issue with the discrete angle finite volume radiation solver sometimes called the "star pattern".  It is due to low angular resolution in the radiation solver.  Read up on the parameters on the RADI line.  I would use something like 400 solid angles (default is 105).  But you may need more.  You can adjust the time step parameters so as to keep the workload the same as the 105 angle case.  An example we use in the FDS validation guide to get very good agreement with heat flux away from the flame is given here:


On Mon, Feb 12, 2018 at 9:15 PM, Priscilla Mariani <priscill...@arup.com> wrote:
Hi,

I intend to simulate a burning timber facade element with a certain net heat flux set on its burning surface to see the effect on its adjacent timber facade elements, whether it radiates enough to cause piloted ignition on the adjacent facade when set at a certain distance.
However, we can't seem to get the expected level of radiative heat flux produced by the burning surface. 
The HRRPUA is set at 110kW/m2 and the radiactive fraction is re-stated as 35%, so we are expecting to see ~38.5kW/m2 radiative heat flux on the burning surface, but the boundary output on it only shows around 15-25kW/m2 radiative heat flux on average with some little spots of 30kW/m2. The incident heat flux on the adjacent facade has thus become too small from what I expected at the set distance.
Also, we expected the radiative heat flux to increase as we go up the facade element, but it gives out a patchy pattern instead of the expected gradient.

I'm trying to understand what is going on and how I can aptly simulate the 35% radiative heat flux.
Could anyone please help?
The input file is attached.

On another note, if anyone has any suggestion on the radiative fraction and soot yield for a pine-type wood, please let me know. 
I couldn't find these from SFPE so I'm just using red oak soot yield and the FDS default radiative fraction, not sure if it's representative enough.
Thank you! :)

--
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.
SMUX_fac_01g.fds

Randy McDermott

unread,
Feb 19, 2018, 5:55:59 AM2/19/18
to FDS and Smokeview Discussions
The inputs look consistent with the example.  Are you getting the correct global radiative fraction?  Plot the Q_RADI column divided by the HRR column in your _hrr file.  It should be 0.35.

To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+unsubscribe@googlegroups.com.

To post to this group, send email to fds...@googlegroups.com.

dr_jfloyd

unread,
Feb 19, 2018, 7:53:31 AM2/19/18
to FDS and Smokeview Discussions
The burner having 110 kW/m^2 does not imply that you should expect to see 38 kW/m^2 in radiative flux. The radiative flux is emitted from the volume of the flame. The surface area of this volume is larger than the burning surface.
Reply all
Reply to author
Forward
0 new messages