reactions stop immediately and furniture disappear

[GoogleCodeExporter]

Please complete the following lines...

Application Version: 5.2.0 Serial
SVN Revision Number:2087
Compile Date: 30 July 2008
Operating System: Win XP

Describe details of the issue below:
I want to simulate a couch fire so I took the example file couch.fds and 
changed only the material properties from the foam, introducing a two step 
chemical reaction.
Running the simulation, at time step 7700 the temperature goes to 0°C, the 
reactions stop immediately and within the next two frames the complete 
furniture had disappeared. 
When I’m running the simulation with the option BURN_AWAY=.FALSE. the 
problems are the same, only the furniture doesn’t disappear.
When I’m running a combustion simulation of a simple box having the same 
properties like the foam I used for simulating the couch the FDS works 
without problems.
Thanks for help!
Erwin

Original issue reported on code.google.com by erw...@virgilio.it on 8 Oct 2008 at 4:13

Attachments:

• couch22.fds
• couch.out
• PUR16.fds

[GoogleCodeExporter]

Try the case with the latest version, 5.2.1, then report back.

Original comment by mcgra...@gmail.com on 8 Oct 2008 at 4:17

• Changed state: MoreInfo

[GoogleCodeExporter]

I tried with version 5.2.1! Same results and same problem, but not more at time 
step 
7700 but at time step 9500.

Original comment by erw...@virgilio.it on 8 Oct 2008 at 5:03

[GoogleCodeExporter]

I've tried running both cases with debug, check all, and fpe trapping.  Both 
have 
gone beyond the failure point indicated above without any sign of instability.

Original comment by drjfloyd on 8 Oct 2008 at 7:19

[GoogleCodeExporter]

Thanks -- I'll run the case as well on my linux machines.

Original comment by mcgra...@gmail.com on 8 Oct 2008 at 8:13

• Changed state: Started

[GoogleCodeExporter]

Ran latest source without any debug and the case runs, but at ~5100 iterations 
CFL 
goes to zero and DIV to 1E5.  

Original comment by drjfloyd on 8 Oct 2008 at 8:35

• Added labels: Application-FDS

[GoogleCodeExporter]

Simo left debugging printout in wall.f90. My case failed at about 214 s, at 
which 
point the volume of a solid cell goes to zero and it is not being trapped. The 
block 
of code looks like:

      IF (VOLSUM > 0._EB) THEN
         K_S(I) = K_S(I)/VOLSUM
         IF (I.EQ.1) E_WALL(IW) = E_WALL(IW)/VOLSUM
      ELSE
         write(*,*) I
      ENDIF

So I see the I's getting printed out, something I assume Simo meant as a way of 
saying that something was wrong. I'll see if I can figure out how to fix.

Original comment by mcgra...@gmail.com on 8 Oct 2008 at 8:52

[GoogleCodeExporter]

This case now seems to run successfully.  Some change(s) made in recent months 
have
fixed the problem.

Original comment by drjfloyd on 9 Jun 2009 at 7:54

• Changed state: Verified