wildland-urban-interface FDS

415 views
Skip to first unread message

berto

unread,
Sep 30, 2008, 8:22:42 AM9/30/08
to FDS and Smokeview Discussions
is the vegetative fuels algorithm available to download?

im interested in using FDS for a potential outdoor tree-to-tree fire
spread model, is it usable in this way? what are the limitations that
the WFDS picks up on?

am i right in saying that the WFDS model only applies to douglas fir
trees?

Ruddy

unread,
Sep 30, 2008, 9:42:44 AM9/30/08
to FDS and Smokeview Discussions
WFDS is the extension of FDS to fire spread and wind through
vegetation. Submodels for drag on the wind flow and thermal
degradation of the vegetation (drying and volatilization) are in place
and are being tested.

When FDS is used for applications involving vegetation users are
encouraged to refer to "WFDS" in the discussion and issue posts in
order to streamline a response.

Tree-to-tree fire spread is handled by WFDS. The user needs to supply
information on the thermally thin vegetation within a "tree" via &PART
namelist (moisture, surface-to-volume ratio, density, initial
temperature, char fraction). The bulk density of the "tree" is also
needed. Any vegetation for which these properties are known can be
included. Whether or not these parameters are enough to adequately
capture fire behavior over a range of vegetative types is not known
conclusively.

Vegetation is represented in Smokeview as particles. These particles
(one per grid cell) are arranged in bulk to form a "tree", "bush", or
"grass". These bulk vegetative volumes can be a cone, cylinder,
frustum, or rectangle. The user gives the dimensions of these shapes
in a &TREE namelist. Alternatively, the user can build up bulk
vegetation cell-by-cell using the rectangular &TREE. Some people have
made separate programs to do this.

A WFDS user guide is under construction.

The FDS source code on the repository contains all the up to date
capabilities for WFDS.

A web page with sample input files and the most up-to-date executables
is
http://www2.bfrl.nist.gov/userpages/wmell/nist_wui_models.html

A web page with some examples of WFDS applications in our wildland-
urban interface (WUI) project is at
http://www2.bfrl.nist.gov/userpages/wmell/public.html

WFDS has been applied successfully to experimental grassland fires and
individual tree burns. It has shows promise for forest stands and fire
spread over terrain. One of the biggest challenges is validation,
since well characterized, repeatable, experiments of large wildland or
WUI fires are hard to come by.

berto

unread,
Sep 30, 2008, 11:05:46 AM9/30/08
to FDS and Smokeview Discussions
thanks Ruddy, thats really helpful. the first link is very useful
indeed, surprising i couldnt find it before.
so i simply download the seperate fuel element and boundary fuel
models into the 'Program Files / NIST / FDS 'working directory, begin
writing input files as illustrated and invoke the program in command
prompt using WFDS.....etc.

no doubt ill be in touch when i have had a chance to get my head
around the sample input files.
> ishttp://www2.bfrl.nist.gov/userpages/wmell/nist_wui_models.html
>
> A web page with some examples of WFDS applications in our wildland-
> urban interface (WUI) project is athttp://www2.bfrl.nist.gov/userpages/wmell/public.html
>
> WFDS has been applied successfully to experimental grassland fires and
> individual tree burns. It has shows promise for forest stands and fire
> spread over terrain. One of the biggest challenges is validation,
> since well characterized, repeatable, experiments of large wildland or
> WUI fires are hard to come by.
>
> On Sep 30, 8:22 am, berto <bert...@gmail.com> wrote:
>
>
>
> > is the vegetative fuels algorithm available to download?
>
> > im interested in using FDS for a potential outdoor tree-to-tree fire
> > spread model, is it usable in this way? what are the limitations that
> > the WFDS picks up on?
>
> > am i right in saying that the WFDS model only applies to douglas fir
> > trees?- Hide quoted text -
>
> - Show quoted text -

Ruddy

unread,
Sep 30, 2008, 11:22:00 AM9/30/08
to FDS and Smokeview Discussions
Your welcome Berto,

Yes, proceed as you described. But note, that the boundary fuel model
for firespread through surface vegetation is based on FDS version 4
and so is no longer supported. However, you should be able to do much
of what you want with the fuel element model. Although it needs to be
more fully tested at coarse resolution the fuel element approach is
producing reasonable firespread results for continuous surface fuels
at grid cell sizes up to 1 m.

berto

unread,
Oct 1, 2008, 5:02:24 AM10/1/08
to FDS and Smokeview Discussions
ok 2 fresh questions relating to this; can WFDS account for fire brand
creation? and have you guys extended your data set to additional tree
species? for example a palm tree?
> > > - Show quoted text -- Hide quoted text -

berto

unread,
Oct 1, 2008, 8:45:07 AM10/1/08
to FDS and Smokeview Discussions
would it be possible to define particles as brands or embers of a
typical size and weight? when particles collide with solid objects
will they transfer heat as well as velocity?

Ruddy

unread,
Oct 1, 2008, 10:38:16 AM10/1/08
to FDS and Smokeview Discussions
WFDS does not presently account for firebrand production, transport,
or ignition via firebrands. Over the next year we are extending our
experimental effort (bench scale, full scale, and field) on
firebrands. One goal of this work is to provide data for firebrand
model development and testing.

Currently the vegetation model is based on the assumption that
firespread is dominated by the fine (thermally thin) vegetation. The
thermally thin vegetation is described in the WFDS input file by the
following: mass of moisture divided by dry mass, surface-to-volume
ratio 1/m, density of dry vegetation kg/m^3, char fraction, drag
coefficient, initial temperature. The bulk density (dry mass of fine
fuels occupying the bulk volume describing the vegetation, e.g. tree
crown) is also needed. See my reply regarding this dated Sept. 30 9:42
am.

The specific heat of the vegetation and its dependence on temperature
follows an expression given in Ritchie et al., "The effects of sample
size on the heat release of charring materials," Fire Safety Science -
Proceeding of the Fifth Intn'l Symp., 177-188, 1997. As is usually
done in FDS, the user can supply the the stoichiometry for the gas
phase chemical reaction and the heat of combustion. A number of
thermal degradation models for the vegetation are currently being
being tested so this can not be modified by the user. The thermal
degradation model currently in WFDS is described in the following
paper on grassland fires
http://www2.bfrl.nist.gov/userpages/wmell/WOW/PAPERS_MINE/2007/mell_wfds_augrassfire_ijwf07_accessory.pdf

So at this point, if a user wants to model a given vegetation (such at
a palm tree, as you asked). They would need to determine the above
listed characteristics of the thermally thin vegetation to use in the
&PART namelist. Then, in the &TREE namlist, they would need to define
the bulk geometry or shape that the themally thin vegetation
occupies.

If the grid resolution allows it, thermally thick vegetation (such as
a tree trunk) could be included following FDS procedures for a
flammable solid.

Ruddy

Ruddy

unread,
Oct 1, 2008, 10:39:31 AM10/1/08
to FDS and Smokeview Discussions
As described in my previous response WFDS does not currently handle
firebrands. This will be added over the next year or so as we build up
our experimental database.

Ruddy

berto

unread,
Oct 1, 2008, 11:03:53 AM10/1/08
to FDS and Smokeview Discussions
when i test your sample input file i get back astack overflow error
severe <170>. please advise

Ruddy

unread,
Oct 1, 2008, 12:32:16 PM10/1/08
to FDS and Smokeview Discussions
Please provide more information:
1) What input file did you use?
2) What wfds executable did you download and run (what is its name and
where did you obtain it)
3) How much memory does your computer have?

Ruddy

berto

unread,
Oct 2, 2008, 5:27:20 AM10/2/08
to FDS and Smokeview Discussions
the first input file for the single tree and also the wfds32.exe file
from the link above in this thread. the machine im trying it with has
2GB RAM

berto

unread,
Oct 2, 2008, 5:36:42 AM10/2/08
to FDS and Smokeview Discussions
i have it working now with reduced grid cell size.
thanks

On Oct 1, 5:32 pm, Ruddy <ru...@nist.gov> wrote:

Ruddy

unread,
Oct 2, 2008, 8:58:11 AM10/2/08
to FDS and Smokeview Discussions
Thanks for checking this out, I did not compile the 32bit serial
version properly. I've put the corrected version on the wfds web page.
When you have a chance please try it out with the original input file.

Also, is the future, for specific questions regarding wfds or fds
runtime problems please submit the problem to the issue tracker. It
helps with our bookkeeping of specific code problems versus more
general considerations.

Thanks, Ruddy

berto

unread,
Oct 2, 2008, 11:16:48 AM10/2/08
to FDS and Smokeview Discussions
aha. i will try downloading again thanks.

btw, what are the FUEL_GEOM shape names that wfds recognises? i notice
it doesnt recognise 'sphere'.

also, what is the characteritsic VEG_SV? and what is set to make sure
the truck doesnt burn? how can i aloow the trunk to burn?

Ruddy

unread,
Oct 2, 2008, 1:53:42 PM10/2/08
to FDS and Smokeview Discussions
FUEL_GEOM defines, in the &TREE namelist, the shape, in-bulk or
macroscopic, of the thermally thin vegetation (e.g., pine needles).
The current choices are (there is not a choice for a sphere at this
point):

FUEL_GEOM='CYLINDER' which requires CROWN_WIDTH (diameter of
cylinder), CROWN_BASE_HEIGHT (height of bottom of cylinder above the Z
value given in the &TREE), TREE_HEIGHT (height of the top of the
cylinder).

FUEL_GEOM='CONE' which requires CROWN_WIDTH (diameter of the base of
the cone), CROWN_BASE_HEIGHT (height of bottom of cylinder above the Z
value given in the &TREE), TREE_HEIGHT (height of the top of the
cylinder).

FUEL_GEOM='FRUSTUM' (i.e., a clipped cone) which requires
CROWN_WIDTH_BOTTOM (diameter of frustum bottom), CROWN_WIDTH_TOP
(diameter of frustum top), CROWN_BASE_HEIGHT, TREE_HEIGHT

FUEL_GEOM='RECTANGLE' a rectangular volume requires the spatial extent
of the rectangle to be defined via XB=xmin,xmax,ymin,ymax,zmin,zmax
which is identical to defining the extent of an obstacle in FDS.

The surface-to-volume ratio of the thermally thin vegetation in define
via VEG_SV. Typical values range from 4000 1/m for pine needles to
10,000 1/m for grasses.

No heat transfer calculation to the tree trunk or stem can will occur
if VEG_STEM=.TRUE. is put in the &PART namelist. The tree will still
present an obstacle to the flow. This requires that the user put in
valid values for VEG_SV for the tree trunk. I've put in VEG_STEM in
the first two sample input files on the wfds web page
http://www2.bfrl.nist.gov/userpages/wmell/nist_wui_models.html

In order to model the burning of a tree trunk set the trunk up as an
obstacle and with surface and material properties. See the section on
materials in the FDS manual.

Ruddy
Reply all
Reply to author
Forward
0 new messages