[fds-smv] oxygen concentration and HRR

[Ricky]

Hello everyone,


I am simulating a fire in an industrial hall (35x15x6 m3) with all the
openings closed. The walls and ceiling of the hall are non
combustible. There is an initial fire source on the floor and fire can
spread to other areas with activation temperature. I have some oxygen
sensors (devices) placed at different locations in the hall. The fire
spreads to floor area with 10 min. and goes out due to lack of
oxygen, as shown by oxygen devices which show zero oxygen
concentration within 10 min. Now I am trying to find a match between
the maximum heat release possible with the available oxygen and that
yielded by simulation. Here is how I calculated it:

Vol. of air = 35x15x6 = 3150 m3
Density of air = 1.18 kg/m3
==> Mass of air = 3150x1.18 = 3717 kg
==> Mass of oxygen = 0.23x3717 = 854.9 kg
EPUMO2 = 13100 kJ/kg
==> Maximum heat release possible = 13100 x 854.9 = 11199321 kJ
==> Total heat released from simulation = 3703156.5 kJ
==> % of maximum heat released in simulation = 3703156.5/11199321 =
33%

Now my questions are:
1. Is the above calculation correct/reasonable?

2. As seen here, only 33% of possible heat release with the available
oxygen is released by simulation, whereas the oxygen devices show zero
oxygen concentration after 10 minutes. What is the reason of this
discrepancy?

3. Is it possible that the oxygen devices can produce wrong results? I
implemented the devices as follows:

&SPEC ID='OXYGEN' /
&DEVC XYZ = 2.80,7.8,1.0, QUANTITY='OXYGEN', ID='OXY1' /
&DEVC XYZ = 2.80,2.6,1.0, QUANTITY='OXYGEN', ID='OXY2' /

Thank you all in advance. I would like to know if anyone else has
faced the same problem.



Best regards,
Ricky


--
You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.
To post to this group, send email to fds...@googlegroups.com.
To unsubscribe from this group, send email to fds-smv+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/fds-smv?hl=en.

[dr_jfloyd]

If you are tracking species with the mixture fraction, you do not need
the &SPED ID='OXYGEN'
Your oxygen devices may not represent the average of the oxygen
everywhere in your domain.
Do you have internal OBST? In which case you need to account for the
air volume occupied by them./=
Do you have any OPEN boundaries? If so then FDS does not compute a
pressure rise unless you use ZONE.

Turn on the MASS_FILE output on the DUMP line and rerun your case.
The chid_mass.csv file will contain the mass of all gasses in your
domain as a function of time.

[Ricky]

Thank you dr_jfloyd for your reply.

No, I don't have any internal OBST. So, the volume of air is correctly
calculated.

I have OPEN boundaries at the computational domain extremities that
are approx. 5 m away from the above mentioned closed hall. So, that
makes the size of my computational domain equal to 40 x 20 x 12 m3.
Will this make any difference? Is there any method to output the
global pressure increase inside this closed hall wrt. to time?

About this mass.csv file, will it output the global value of gas
species (oxygen, co2 etc.) inside the domain wrt. time?

In the mean time, I will re-run the simulation and will report the
result if it is different.

I hope to hear from you again. Thank you.


regards,
Ricky

[Ricky]

Forgot to mention one more thing, if I dont specify &SPED ID='OXYGEN',
then I get the error message "ERROR: DEVC QUANTITY OXYGEN not found".

[dr_jfloyd]

If you have an open boundary and you have not specified a ZONE for
your enclosure, then FDS will not compute a pressure rise in the
enclosure. If pressure is always 1 atm, as the temperature in your
compartment rises, you will in effect be losing mass. Either close
the open boundaries or define a ZONE for your compartment.

QUANTITY='OXYGEN' is not a valid QUANTITY.


QUANTITY='MASS FRACTION',SPEC_ID='OXYGEN'

[Ricky]

Thank you for your reply.

Ok, I will define a pressure zone for the volume enclosed within solid
boundaries. But, in another simulation, I have to open door and
windows at a certain temperature. So, will this pressure zone affect
in that case? Will activating a opening abutting against the boundary
of pressure zone break this pressure zone? Or I have to define
something else to break it?

regards,
Ricky

[dr_jfloyd]

Read the User's Guide section on pressure zones.

[chemboy]

Hi
it is actually default value of EPMUO2 which is 13100.
Do you think it is in accordance with the material that you are
burning in the industrial hall.
secondly activation temperature is used for activation of sprinkler or
heat detector. how activation temperature is used to spread the fire
in the industrial area.
I think heat of combustion simulated by FDS may depend upon the
material properties that you are burning.
but you stick to the form and may be some one else with more good
knowledge can help you.
regards
asim

[Ricky]

Hello everyone,


Defining a pressure zone solved my problem concerning discrepancy
between oxygen consumed and heat released. But now I have negative
convective heat loss and energy does not seem to be conserved (sum of
conductive and convective losses are not equal to HRR). what could be
the reason for this? And temperature inside the hall now are quite
higher than before.

dr_jfloyd, could you please elaborate me a bit, how oxygen mass is
lost when I do not define a pressure zone and pressure inside the room
is always atmospheric(case before), as you said above?

thank you.

regards,
Ricky

On May 12, 9:50 am, chemboy <asim...@gmail.com> wrote:
> Hi
> it is actually default value of EPMUO2 which is 13100.
> Do you think it is in accordance with the material that you are
> burning in the industrial hall.
> secondly activation temperature is used for activation of sprinkler or
> heat detector. how activation temperature is used to spread the fire
> in the industrial area.
> I think heat of combustion simulated by FDS may depend upon the
> material properties that you are burning.
> but you stick to the form and may be some one else with more good
> knowledge can help you.
> regards
> asim
>
> On 11 Maj, 10:30, Ricky <alan.ri...@ymail.com> wrote:
>
>
>
> > Hello everyone,
>
> > I am simulating a fire in an industrial hall (35x15x6 m3) with all the
> > openings closed. The walls and ceiling of the hall are non
> > combustible. There is an initial fire source on the floor and fire can
> > spread to other areas with activation temperature. I have someoxygen
> > sensors (devices) placed at different locations in the hall. The fire
> > spreads to floor area  with 10 min. and goes out due to lack of
> >oxygen, as shown byoxygendevices which show zerooxygen
> > concentration within 10 min. Now I am trying to find a match between

> > the maximum heat release possible with the availableoxygenand that

> > yielded by simulation. Here is how I calculated it:
>
> > Vol. of air = 35x15x6 = 3150 m3
> > Density of air = 1.18 kg/m3
> > ==> Mass of air = 3150x1.18 = 3717 kg

> > ==> Mass ofoxygen= 0.23x3717 = 854.9 kg

> > EPUMO2 = 13100 kJ/kg
> > ==> Maximum heat release possible = 13100 x 854.9 = 11199321 kJ
> > ==> Total heat released from simulation = 3703156.5 kJ
> > ==> % of maximum heat released in simulation = 3703156.5/11199321 =
> > 33%
>
> > Now my questions are:
> > 1. Is the above calculation correct/reasonable?
>
> > 2. As seen here, only 33% of possible heat release with the available
> >oxygenis released by simulation, whereas theoxygendevices show zero

> >oxygenconcentration after 10 minutes. What is the reason of this
> > discrepancy?
>
> > 3. Is it possible that theoxygendevices can produce wrong results? I

[dr_jfloyd]

A pressure rise in a closed volume results in an increase in
temperature due to the work required to raise the pressure. You must
account for this pressure work when doing an energy balance.

For your other question please carefully review the Technical guide.
If there is an open boundary on a face, then the entire face gets the
Dirichlet boundary condition. As a result there may be a small normal
velocity. One can set VELOCITY_TOLERANCE to reduce this error at the
cost of additional CPU time due to increased pressure iterations.