HRR: Incorrect calculation

[gforney]

Application Version: 5.5.3 serial
SVN Revision Number: 7031
Compile Date: Fri, 29 Oct 2010
Operating System: Windows XP, 32bit

Application Version: 5.5.3 serial
SVN Revision Number: 7031
Compile Date: Fri, 29 Oct 2010
Operating System: Windows 7, 64bit

Application Version: 5.5.3 serial and parallel
SVN Revision Number: 7031
Compile Date: Fri, 29 Oct 2010
Operating System: Linux 

-----------------

After change the FDS Version (5.5.2 -> 5.5.3) I checked my cases without any changes.
In the energy budget (_hrr.vsv - 5.5.3 calculation) I found one interesting phenomenon.
The convective/ radiative loss is very very high. When I compared the output files,
there is no big difference between the results from 5.5.2 and 5.5.3 - only in the conv_loss
calculation. I can reproduce the issue on different OS with the offical version. 

Can you check the source code? I think, there is some mistake in the calculation/ output
without impact on the rest.

Thanks

Gregor

Original issue reported on code.google.com by gregor.jaeger on 2010-11-16 09:55:20

---

• _Attachment: issue_hrr-csv.pdf_

[gforney]

Please post the input file. Thanks.

Original issue reported on code.google.com by mcgratta on 2010-11-16 13:40:15

[gforney]

no problem:
issue-hrr.fds - simple input-file without geometry and fans. 
issue-hrr_hrr.csv - output-file (first steps)

Original issue reported on code.google.com by gregor.jaeger on 2010-11-16 14:17:17

---

• Attachment: issue-hrr.fds
• _Attachment: issue-hrr_hrr.csv_

[gforney]

We corrected a discrepancy between the User's Guide and the code.

http://code.google.com/p/fds-smv/source/detail?r=7031

Original issue reported on code.google.com by drjfloyd on 2010-11-16 14:32:37

[gforney]

I think that this is a bug. The change to the code was to account for radiation heat
transfer through OPEN boundaries. However, the terms QRADIN and QRADOUT do not appear
to be set at OPEN boundaries.

Original issue reported on code.google.com by mcgratta on 2010-11-16 15:44:04

[gforney]

In init.f90

IF (RADIATION) THEN
...
   M%QRADIN(IW)  = M%E_WALL(IW)*SIGMA*TMPA4
   M%QRADOUT(IW) = M%E_WALL(IW)*SIGMA*TMPA4
ENDIF

We are just missing the update to the outgoing radiation in radi

Original issue reported on code.google.com by drjfloyd on 2010-11-16 16:02:45

[gforney]

Fixed.  A simple test case of a hot surface with h_fixed=0 has the sum of CONV_LOSS
and COND_LOSS converge to zero as grid resolution is increased.

Original issue reported on code.google.com by drjfloyd on 2010-11-17 16:19:21

• Labels added: OpSys-All, Application-FDS

[gforney]

Thanks.

Original issue reported on code.google.com by mcgratta on 2010-11-17 16:53:54

[gforney]

The fix caused a problem. WALL_CELL(IW)%ILW(:,:) is not allocated at OPEN boundaries.
I don't know how to save radiation data at OPEN boundaries. I will cc Simo.

Original issue reported on code.google.com by mcgratta on 2010-11-17 18:53:53

[gforney]

(No text was entered with this change)

Original issue reported on code.google.com by drjfloyd on 2010-11-17 19:10:40

[gforney]

Thanks, I read with interest.

When you fixed the problem, can you implement a test case in the verification folder
"energy_budget"?

Original issue reported on code.google.com by gregor.jaeger on 2010-11-17 19:41:05

[gforney]

Simo -- we need to store the radiation flux going out OPEN boundaries. I would rather
not allocate WALL_CELL(IW)%ILW(:,:) at all OPEN boundaries because it takes up alot
of RAM. Could we create a single IW array to save the flux that would be computed during
the main radiation loop?

Original issue reported on code.google.com by mcgratta on 2010-11-17 19:55:11

[gforney]

I have now added the computation of QRADIN for OPEN boundaries. QRADOUT = sigma*TMPA^4.
Enclosed is the plot of HRR and COND_LOSS + CONV_LOSS in the simulation file above
(mesh resolution reduced for faster testing).

Original issue reported on code.google.com by shostikk on 2010-11-22 14:01:40

---

• _Attachment: issue-hrr_hrr.csv.pdf_

[gforney]

The changes were commited at revision: 7182  

Original issue reported on code.google.com by shostikk on 2010-11-22 14:02:46

[gforney]

Between FDS 5.5.3 and FDS 6-beta, I tested the case with revision 7985. It seems to
me, that it run fine. I will test the case with the offical FDS 6-beta version.

Original issue reported on code.google.com by gregor.jaeger on 2011-03-27 18:47:07

[gforney]

(No text was entered with this change)

Original issue reported on code.google.com by mcgratta on 2012-07-21 19:27:33