_hrr.csv energie balance

[gforney]

Application Version: FDS 5.4.3 32bit
SVN Revision Number: 5210
Compile Date: 03 Dec 2009
Operating System: Win XP

Describe details of the issue below:

I have a question concerning the hrr.csv output file.

CONV_LOSS + COND_LOSS = HRR

I am doing this with a quite simple test case (adiabatic surfaces). See 
attached input file.

See attached PDF:
Even though there are only adiabatic surfaces, there is a significant 
amount of cond_losses.
Conv_losses it self already is slightly higher than hrr.
The sum of cond and conv losses is ~ 100 kW > hrr.

Right now I don’t know what I am getting wrong here...

Thanks for your help.

Boris

Original issue reported on code.google.com by Info@F-Sim.de on 2010-02-25 17:58:48

---

• _Attachment: test_adiabat_hrr.pdf_
• _Attachment: test_adiabat.fds_
• _Attachment: test_adiabat_s0000.png
  _
• _Attachment: test_adiabat_hrr.csv_

[gforney]

I'll take a look at it.

Original issue reported on code.google.com by mcgratta on 2010-02-25 22:36:52

[gforney]

Not all of your surfaces are adiabatic.  SURF lines do not inherit any attributes
from earlier SURF lines, thus, SURF_ID='FIRE' is Fixed to TMPA.

Original issue reported on code.google.com by drjfloyd on 2010-02-26 12:42:21

[gforney]

Ah ha! That's what it is. I couldn't figure this one out. I also wondered, but 
didn't have time last night, if some of the extra convective heat could be coming in

through the OPEN boundary because of the averaging procedure to get RHO at the 
boundary. It is a coarse grid, and these kinds of spurious effects are worst on a 
coarse grid.

In any case, does the ambient burner temperature explain all the anomolies, or just

the non-zero conduction into the walls?

Original issue reported on code.google.com by mcgratta on 2010-02-26 12:51:31

[gforney]

I haven't run the case, I just had a feeling that Boris had more than one SURF line.

Original issue reported on code.google.com by drjfloyd on 2010-02-26 12:58:14

[gforney]

Kevin,

"It is a coarse grid"
I did a much bigger simulation (1.600 m², 12,5 cm cells) with a much finer grid and

also spotted significant differences...

Original issue reported on code.google.com by Info@F-Sim.de on 2010-02-26 13:10:01

[gforney]

Attached you find the same example without adiabatic surfaces.

Original issue reported on code.google.com by Info@F-Sim.de on 2010-02-26 13:15:37

---

• _Attachment: test_brick_hrr.csv.csv.pdf_

[gforney]

And one more thing of interest.
I attached the same case, done with FDS 5.4.2.

Original issue reported on code.google.com by Info@F-Sim.de on 2010-02-26 13:20:18

---

• _Attachment: test_adiabat_hrr.csv.csv.pdf_

[gforney]

Thanks. Setting ADIABATIC=.TRUE. on the burner SURF line results in the COND_LOSS 
returning to zero. However, CONV_LOSS (convective loss due to the flow in and out)

is still higher than HRR. This might be due to the OPEN boundary condition, as I 
indicated above.

Original issue reported on code.google.com by mcgratta on 2010-02-26 13:36:54

[gforney]

There is definitely something different between 5.4.2 and 5.4.3. The attached case 
is even simpler than the original because it eliminates radiation completely. I'll

try to track it down.

Original issue reported on code.google.com by mcgratta on 2010-02-26 16:16:16

---

• _Attachment: test_adiabat.fds_

[gforney]

SVN 5052 works properly. SVN 5056 does not, or let's just say that the difference in

results occurs between these two revisions. Specifically, read.f90 and func.f90 are

the two files that cause the change.

Original issue reported on code.google.com by mcgratta on 2010-02-26 17:44:24

[gforney]

I've posted new source that should fix this problem.  The issue was related to how we
were dealing with the enthalpy of water vapor (which could arise either by combustion
or by evaporation of a water drop).  Running the case with no radiation and all
surfaces adiabatic (including the burner), there is no longer a significant error in
the energy balance.

Original issue reported on code.google.com by drjfloyd on 2010-02-26 21:42:46

• Labels added: Priority-Critical, OpSys-All, Application-FDS
• Labels removed: Priority-Medium

[gforney]

Also, as far as I can tell, the only real impact of this error was when going to
compute the CONV_LOSS

Original issue reported on code.google.com by drjfloyd on 2010-02-26 21:44:14

[gforney]

Thanks for your fast help.

A question: is it possible to include a simple verification test in the verification
guide that confirms the energy balance in the hrr.csv?

Original issue reported on code.google.com by Info@F-Sim.de on 2010-02-26 22:16:11

[gforney]

Yes, your case. I am going to add several versions of it, including the simple case

with adiabatic walls.

Original issue reported on code.google.com by mcgratta on 2010-02-26 22:26:12

[gforney]

Thanks...

Original issue reported on code.google.com by Info@F-Sim.de on 2010-02-27 09:54:58

[gforney]

I added two new verification cases based on the original file posted above. One with

adiabatic walls, the other with cold walls, in the repository folder 

Verification/Energy_Budget

Original issue reported on code.google.com by mcgratta on 2010-02-27 20:22:36

[gforney]

The water_evaporation verification case is not working properly (or at least as it 
was). 

Original issue reported on code.google.com by mcgratta on 2010-03-02 14:55:29

[gforney]

fixed

Original issue reported on code.google.com by drjfloyd on 2010-03-02 20:09:52

[gforney]

case "test_adiabat.fds" verified with FDS 5.5.2 (SVN 6706)

Original issue reported on code.google.com by gregor.jaeger on 2010-09-18 08:50:34

[gforney]

Thanks

Original issue reported on code.google.com by drjfloyd on 2010-09-18 14:45:33

[gforney]

I am just marking this case as "Verified."

Original issue reported on code.google.com by mcgratta on 2010-09-20 12:53:01