Soot Yield for Polyurethane for input to FDS.

[Cavalier]

Dear FDS users,

I've been using the Database 4 to get the soot yield for Polyurethane
( Which is referenced from NFPA - Babrauskas-App A). The figure is
quiet high , 0.1 kgsoot/kgspecimen. This has always been a major bug
bear for me when i model seats and furnitures burning because the
visibility criteria is always quiet hard to meet. Recently i came
across another article on SFPE - Mulholland, Table 2-13.1 where the
the value for the smoke conversion factor for Polyurethane- Flexible-
Flaming is <0.01 kg/kg. This value is about 10 times less than the
quoted soot yield. I understand that Table 3.5 - Principles of smoke
management, JohnH.Klote also specifies that soot yield for
Polyurethane is about 0.1.

I'm in the midst of modeling the furniture- fixed seating in airport
and am in a quandary on which value to use for the soot yield.Would
appreciate if anyone with such experience to advise.

Thanks,
Victor

[BlairS]

Hi Victor,
I suspect a soot yield of 0.01 would not be a conservative design
figure. This is the low end of a range of values for flaming
combustion of a sample with area 0.005 m^2 (i.e. small scale). The
range goes up to 0.035 but you haven't quoted that - why? What makes
you think the results will scale up? The 'database' file for FDS4 was
meant as a starting point to test the model's features, I recall the
user's guide explicitly stated that the user should research and use
his/her own values (the fact that nobody did this is the reason the
database is no longer provided).

Under well-ventilated flaming conditions with this one specific type
of polyurethane foam in the cone calorimeter, you may get a soot yield
in this range, however during the growth phase, in a real fire, or
with a different incident heat flux, the yield could be much higher.
Under-ventilated conditions will affect the amount of soot produced.
Even the foam at the corner of an armrest could yield more soot than
the top of the seat. Sprinklers will also have an effect.

The SFPE Handbook of Fire Protection Engineering (3rd Edn, Table
3-4.14) gives flexible polyurethane soot yields up to 0.227 kg/kg fuel
burnt. I'm not saying this is the correct value to use for your
situation either, but it is up to you as designer to justify the soot
yield used. If I were reviewing your design I would question use of a
yield of 0.035. Unless you have test information for the specific
fuel package, which includes the exact combination of timber, foam and
fabric to be used in that area forever, I'd suggest being conservative
is the only way to go. You may wish to consider using a high
percentile of a range of values from Table 3-4.14.

The visibility criteria are hard to meet for several reasons. We are
dealing with people's lives here (I don't mean to preach, but it's
easy to forget sometimes when you're just looking at an empty model),
and the test data on how the people using your building will respond
physiologically or psychologically to different types and quantities
of smoke is very limited.

Please try to avoid the temptation to reverse engineer input
parameters to achieve the result you think you should get - otherwise
why do we do modelling at all?

Regards,
Blair Stratton
Beca

[Jianping]

I agree with Blair that for Engineering design one should be on the
safe side.

Here I just want to add a comment, in FDS the soot mass fraction is
assumed to be dependent on only mixture fraction via modifying the
state relationship by including CO, soot using constant yields. This
is a very crude approximation, and may not reflect what happens in
reality. We have done simulations for small pool fires and found that
the soot volume fraction predicted is rather different from
experimental data, and we had to tune the constant soot yield (soot
volume fraction is proportional to soot yield) to get reasonable
agreement.

The point I am trying to make here is that the choice of the constant
soot yield depends on in addition to fuel types many other factors
(such as fire scenarios, prediction of the flow field), and I think
one has to keep this in mind for practical designs and be cautious
when interpreting the simulations results.

Jianping

[Chris]

In my opinion a standard soot yield value for some different materials
should be available for every advanced fire modelling user. Otherwise
you come out for the same case with completely different results. The
effort to get a soot yield value for different materials is too high
for a single person or a small engineering consultancy and the soot
yield is dependent on a variety of factors as written above. That
would not be a high-risk problem but the soot yield is a very critical
input value for the software and it would make the people feel more
comfortable by using a standard soot yield value.

> > > Victor- Zitierten Text ausblenden -
>
> - Zitierten Text anzeigen -

[BlairS]

Chris,
A single approved set of material input parameters would certainly
make things easier for achieving building approval but I don't believe
this is the best way to go. Robust performance based designs need
good justification for the models used, parameters input, assumptions
made and how the results are interpreted. If some body simply deemed
that "from now on all models must be run with a soot yield of 0.1" or
whatever, this would be a step back towards prescriptive design, and
may be conservative in some cases and not in others.

The info is generally available with very little searching required,
even for small consultants. Members of this group generally seem
happy to point users in the right direction if they get stuck.

Regards,
Blair

> > - Zitierten Text anzeigen -- Hide quoted text -
>
> - Show quoted text -

[Jason]

In my opinion, the value of soot production is an issue related to the
design fire, not FDS.
Since FDS is just a computer modelling tool to process your input.
I believe a standard guideline for choosing the soot rate will be
helpful, but this guideline should be based on comprehensive research
work.

Jason
PB

[Chris]

I certainly agree that the input values have to be scientific verified
the best it can be done. Nevertheless I for myself would be a bit more
confident with some standard range values. I would say these standard
input-values for building design could even be sold by NIST as I know
there is a lot of work behind.

[Kevin]

NIST is an agency of the US Department of Commerce. It is not a
business. All of our research is in the public domain, including our
fire models. We don't sell data.

> > PB- Hide quoted text -

[Cavalier]

Dear All,

Thanks for the inputs on the Polyurethane Soot Yield. I do agree that
we can't just assume a standard
number and assume it as the silver bullet to solve the soot yield
problem. I'll be approaching
this problem by looking at the percentage of Polyurethane as part of
the chair and multiplying
the fraction by the soot yield. A good article that i found on the net
about the percentage of
Polyurethane as part of the chair with the picture and dimension of
the chair given is in the
article below.
www.abcb.gov.au/index.cfm?objectid=0273B4F4-F365-B0BB-E75DE75958B75232
-

Thank You all for your help.
Victor
Singapore

> > - Show quoted text -- Hide quoted text -