pyrolysis and the solid phase model for wood

[vanessaL]

Bryan asked me to post my question in the group, and here is my first
post:

I badly need a good model for gypsumboard, solid wood, plywood, etc
for my structural model and am therefore very interested in the FDS
pyrolysis and the solid phase model. In particular I would be grateful
if you could let me know if the FDS pyrolysis model is good at
modeling the wood? how much work has been done for wood and wood
products material properties in FDS?

Many thanks!

[Kevin]

First thing to consider is that most pyrolysis routines used in fire
models only consider the thermal behavior of these materials, not the
mechanical. For example, wood pyrolysis can be modeled as a series of
reactions that evaporate moisture, convert virgin wood to char, ash,
lignin, etc. Some models are very simple, some complex. Depends on the
application.

Second, the problem faced by fire researchers is not, in my opinion,
the model itself, but rather the material properties. There are models
of thermal decomposition and pyrolysis that capture much of the
underlying physics, but there is no simple way to measure the many
property inputs needed. There have been numerous threads on this
subject.

You need to be more specific about what you mean by "good model". What
physical phenomena do you want the model to represent?

[philio]

Alot of work has been done in FDS for wood and other popular
materials .. like polyurethane pmma pvc, pure liquid fuels etc

But wood I think is one of the trickiest, because it is not
homogenius, and it chars . that doesn't mean that we can derive
certain assumptions about how to model it in fds, but to capture the
exact physics of it would be tricky.

think about this though. If took a volume and mass of Pine wood, and
ignited it in a calorimeter, and determined how much heat it produces,
peak HRR, its required ignition energy etc... we could then take this
and apply it to FDS.

A very advanced research paper by Anna Matala:
Estimation of solid phase reaction parameters for fire simulation :
http://www.sal.hut.fi/Publications/pdf-files/TMAT08.pdf

This may be above and beyond what you are trying to do, AE takes alot
of analysis time.
You may have to look for these other materials such as plywood,

This is for structural application? A burning wall assembly and you
would be measuring internal structure temperatures?

On Feb 3, 3:15 pm, vanessaL <vanessa.lin...@gmail.com> wrote:

[clauten]

This paper is (still) under review:

http://me.berkeley.edu/cpl/publications/2008-01-26_C+F_oxidative_pyrolysis.pdf

but it contains material properties for white pine (thermophysical
properties, reaction kinetics, radiative properties, etc.). The
material properties are extracted from bench-scale pyrolysis
experiments under different atmsopheres by genetic algorithm
optimization using a pyrolysis model called Gpyro, and the derived
properties are compared with literature data.

The focus of this paper is the effect of oxygen concentration on wood
pyrolysis because going from an inert atmosphere to air causes wood's
mass loss rate to double and its surface temperature to increase by
200 deg C due to char oxidation. In a fire you get both -- a nearly
inert environment when a surface is covered by a diffusion flame, but
an oxidative environment elsewhere.

Due to oxidative reactions (both heterogeneous and homogeneous gaseous
reactions occurring inside voids) and other physical phenomena
included in Gpyro but not FDS (radiation heat transfer across pores,
diffusion of oxygen from the freestream into porous media) the
material properties cannot be directly ported to FDS. However, you can
probably get a a good representation of pyrolysis under nitrogen in
FDS by stripping the oxidative reactions out of the reaction
mechanism.

[Rein]

You will find very useful information in this recent PhD thesis on
wood pyrolysis for structural applications:

"In-Depth Temperature Profiles in Pyrolyzing Wood" (Reszka, 2008),
University of Edinburgh.
http://www.era.lib.ed.ac.uk/handle/1842/2602

Note that (as said in the abstract) "the main objective was to
identify the simplest model that can accurately predict temperature
distributions in wood elements exposed to fires". The thesis also
contains one of the most complete sets of measurements of wood
pyrolysis to date.

Let me know if you are interested in contacting Dr Reszka.

Cheers
G.

[Ling Lu]

Thanks to all the replies!

 

I need a temperature profile for the floor/wall assembly as input for the structural model. I have a thermal model in which the material properties change with the temperature. I also have the model from Steven Craft "Predicting the Fire Resistance of Light-Frame Wood Floor Assemblies" which is a heat and mass transfer model, but it is only valid up to the point where the gypsum board falls off. For the structural model, I need a temperature profile for both gypsum board and solid wood joists. I will read through the papers you have recommended and see which is  closest to my interests. Dr Reszka's "simplest model that can accurately predict temperature distributions in wood elements exposed to fires" is very attractive to me.

 

What I was wondering is whether FDS' pyrolysis model with multilayer and mass fraction of different components can do a better job than what I have now. But I realize that Kevin is right, it is rather the material properties than the model that puzzled me.

 

Would you think it useful, as Philio suggested, to do a bit of research on a piece of gypsum board, ignited in a calorimeter, to determine how much heat it produces, peak HRR, required ignition energy etc... then repeat the process with a piece of gypsum board+wood.... and apply the results to FDS or other models?

[bryan...@nist.gov]

Do you think that Dr. Reszka would have good information to add to
this discussion.
Please pass along an invitation to join the group and participate in
this thread.

-Bryan Klein

On Feb 4, 9:58 am, Rein <rei...@gmail.com> wrote:
> You will find very useful information in this recent PhD thesis on
> wood pyrolysis for structural applications:
>
> "In-Depth Temperature Profiles in Pyrolyzing Wood" (Reszka, 2008),

> University of Edinburgh.http://www.era.lib.ed.ac.uk/handle/1842/2602

[Kevin]

There are numerous papers in the fire literature that address these
various issues. No matter how simple or detailed your model, I predict
your stumbling block is going to be the property data. There is no
consensus on how to measure it, or even what to measure.

If you are modeling a floor or wall assembly that is exposed to a
constant heat flux or temperature, you might try gathering bench-scale
data at the same flux. But then keep in mind that you are only
calibrating the model -- you are not predicting anything. That might
be OK, depending on your objective.