Incident radiative heat flux on virtual surfaces or points

78 views
Skip to first unread message

D D

unread,
Jan 20, 2021, 10:09:12 AM1/20/21
to FDS and Smokeview Discussions
I m facing the same concerns as the original poster in this thread


which I haven't been able to fully clear from the writings in it.

What I want to calculate is the incident heat flux (IHF) on virtual surfaces or points, that lie where no obstruction or end of domain exist. One may here say, why would I want to measure the IHF on something that does not exist, that does not absorb or radiates back. Although it may sound like a reasonable question, there are cases where through the development of a project there is the need to exhibit findings and results and discuss over them to assess whether something proposed is feasible on the path of radiation or not.

From the other hand, adding a temporary obstruction to place a device is to me a not so good idea simply because it may affect the fire phenomena taking place.

So, is it possible to calculate the IHF on a point that lies in the gas-phase area of the domain, that is, not lying on a solid surface?

I tried the following found in the FDS UG, 20.10.8

&DEVC ID='AST', XYZ=..., QUANTITY='ADIABATIC SURFACE TEMPERATURE GAS', ... /
&DEVC ID='HTC', XYZ=..., QUANTITY='HEAT TRANSFER COEFFICIENT', INIT_ID='AST' /

changing only the second line as
&DEVC ID='HTC', XYZ=..., QUANTITY='INCIDENT  HEAT FLUX', INIT_ID='AST' /

but does not work. Can somebody provide some information whether this is possible to make it work and if so, how?

Also, in the thread above I read about lagrangian particles. However, I could not find much info. Could this be a possible solution?

Thank you.

Kevin McGrattan

unread,
Jan 20, 2021, 10:33:52 AM1/20/21
to fds...@googlegroups.com
I suggest you use Eq 21.14 of the most recent FDS User's Guide to solve for q_inc. FDS is currently not configured to do what you want.

D D

unread,
Jan 20, 2021, 10:56:33 AM1/20/21
to FDS and Smokeview Discussions
Thank you. Could you please elaborate a bit more about the convective heat transfer coefficient found in the equation? Is it something that the user should calculate by themselves or is it something FDS can provide? I am aware of the 'Heat Transfer Coefficient' quantity. Is this meant as the overall or is this the convective coefficient?

Kevin McGrattan

unread,
Jan 20, 2021, 10:59:22 AM1/20/21
to fds...@googlegroups.com
You should provide the heat transfer coefficient. It is the convective coefficient, or h. Since the actual solid surface does not exist, there is no way for FDS to calculate it. You should provide it.
Reply all
Reply to author
Forward
0 new messages