search

firemodels / fds

Fire Dynamics Simulator
https://pages.nist.gov/fds-smv/
Other
648 stars 618 forks source link

Heat Transfer to a bare Steel - FDS6 vs. FDS5 #2196

Closed gforney closed 9 years ago

gforney commented 9 years ago 
Please complete the following lines...

FDS Version: FDS 6.1.1; MPI Disabled; OpenMP Enabled
SVN Revision Number: 19882
Compile Date: Thu, 10 jul 2014

FDS Version: FDS 5.5.3; MPI Disabled; OpenMP Enabled
SVN Revision Number: 7051
Compile Date: Mo, 01 Nov 2010

Smokeview Version/Revision: 5.6 - Oct 29 2010
Operating System: Windows 7

Describe details of the issue below:
I have just tried to calculate the heat transfer to a bare steel with FDS6 and FDS5.
the difference between the two models is huge. I will not be able to follow this issue.
I also want to create a Post-Flashover situation with the two models, because in my
opinion in such a situation the dependence of the boundaries are weak? Here are the
files some plots.

So any idea?

Original issue reported on code.google.com by abdellaoui.chalid on 2014-09-05 08:42:22

gforney commented 9 years ago 
The FDS 6 results are what I would expect. By definition, the actual steel temperature
will approach the adiabatic surface temperature if the thermal exposure conditions
are quasi-steady. What is the point of working with both models, now that you have
identified a potential bug in the old version? When we change the major version number
of FDS, we fix many bugs and we improve the algorithms, at least we hope too.

Also, why are you using all these extra features on the MISC line? Many of these features
were in testing mode in FDS 5. Even in FDS 6, we keep them around mainly for diagnostic
purposes. Why not just use the defaults whenever possible unless you have a good reason
to use non-standard parameters. This is particularly important if you find unusual
behavior. The FDS validation cases are set up in a way that a typical user would do
it. Thus, we haven't thoroughly V&V'ed some of these less often used parameters.

Original issue reported on code.google.com by mcgratta on 2014-09-05 12:29:44

gforney commented 9 years ago 
What is the point of working with both models:
To compare both of them, normaly there shoud´nt be such a huge difference.
The differences are there not only for the quasi-steady state but also in rising phase
(about 30°C difference). 

Also, why are you using all these extra features on the MISC line? 
which ones? 
in FDS6 I tried to set the same parameter as FDS5. I´ve started a simulation with default
fds6 parameters that I interrepted because of the same huge diefference I remarked.

Original issue reported on code.google.com by abdellaoui.chalid on 2014-09-05 13:13:56

gforney commented 9 years ago 
If you see such a huge difference in the two models, this is a sign of either a bug
in the old code or that your inputs are not the same.  In your case you did not recognize
that the default BACKING in FDS5 is not the same as in FDS6 and the default backing
in FDS5 is inappropriate for your simulation.

Original issue reported on code.google.com by drjfloyd on 2014-09-05 13:38:28

gforney commented 9 years ago 
I forgot to tell you I have run the same case with StarCCM+ without counting for radiation
in the StarCCM+ calculation. The results compare within acceptable range.

Original issue reported on code.google.com by abdellaoui.chalid on 2014-09-05 13:52:51

gforney commented 9 years ago 
in 15 min I will leave the Office. next week I will try to look for the backing.  nice
weekend

Original issue reported on code.google.com by abdellaoui.chalid on 2014-09-05 13:57:13

gforney commented 9 years ago 
In your FDS 5 results, how can a beam immersed in gas with an adiabatic temperature
of 300 C not exceed 220 C?  Zeroth Law of Thermodynamics would say this can't happen,
that heat transfer will occur until the temperature difference is 0.  This is the result
of the FDS 5 default condition of BACKING='VOID' which is inappropriate vs. FDS 6 with
BACKING='EXPOSED' which is the appropriate condition.

Original issue reported on code.google.com by drjfloyd on 2014-09-05 14:02:24

gforney commented 9 years ago 
Hi there I´m back, you´re right about the backing. But if you look the adiabatic surface
temperature you will see a difference of 150 °C to 300 °C. 
So for the same heat release FDS 6 is generating higher temperatures than FDS 5.
The inputs are simple and easy to check. Any idea?

Original issue reported on code.google.com by abdellaoui.chalid on 2014-09-16 07:29:25

gforney commented 9 years ago 
In transitioning from FDS 5 to 6, we fixed many minor bugs related to ADIABATIC surfaces,
back side boundary conditions, and so on. At this point, I am only interested in knowing
if FDS 6 is significantly different than experimental measurements or basic heat conduction
theory. We do not have staff members to devote to examining old code.

Original issue reported on code.google.com by mcgratta on 2014-09-16 12:47:57

gforney commented 9 years ago 
OK! and thanks

Original issue reported on code.google.com by abdellaoui.chalid on 2014-09-16 12:53:35