Large Eddy Simulations of Smoke Movement in Three Dimensions.
Large Eddy Simulations of Smoke Movement in Three
Baum, H. R.; McGrattan, K. B.; Rehm, R. G.
NISTIR 6030; June 1997.
Interscience Communications Ltd.; National Institute of
Standards and Technology; Building Research
Establishment; and Society of Fire Protection Engineers;
Swedish National Testing and Research Institute.
Interflam '96. International Interflam Conference, 7th
Proceedings. March 26-28, 1996, Cambridge, England,
Interscience Communications Ltd., London,England,
Franks, C. A.; Grayson, S., Editors, 189-198 pp, 1996
AND U.S./Japan Government Cooperative Program on
Natural Resources (UJNR). Fire Research and Safety.
13th Joint Panel Meeting. Volume 1. NISTIR 6030. March
13-20, 1996, Gaithersburg, MD, Beall, K. A., Editor,
249-258 pp, 1997 AND European Symposium on Fire Safety
Science, First (1st). ABSTRACTS. Session II. CFD 1.
Proceedings. ETH Institute for Structural Engineering,
Zurich, Switzerland, II-5/75-76, August 21-23, 1995,
Available from: National Technical Information Service
(NTIS), Technology Administration, U.S. Department of
Commerce, Springfield, VA 22161.
1-800-553-6847 or 703-605-6000;
ON BOOK SHELF: TH9112.I57 1996
Order number: PB97-184204
fire safety; smoke measurement; simulation; mathematical
models; equations; smoke transport; high temperature
This paper describes a methodology for simulating the
transport of smoke and hot gases in enclosures. The
approach is based on the use of efficient CFD techniques
and high performance computers to solve a form of the
Navier Stokes equations specialized to the smoke
movement problem. The fire is prescribed in a manner
consistent with a mixture fraction based approach to
combustion, but the combustion phenomena themselves are
not simulated. The mixing and transport of smoke and
hot gases is calculated directly from an approximate
form of the Navier Stokes equations. The computations
are carried out as a three-dimensional time-dependent
process, limited only by the spatial resolution of the
underlying grid. No turbulence models are employed; the
large scale eddies are simulated directly and sub-grid
scale motions are suppressed. Present capabilities
permit a typical residential room or hotel unit to be
simulated at a 3-5 centimeter resolution limit, with
correspondingly coarser resolution for larger spaces.
The enclosure can have any shape made up of rectangular
blocks, and can be multiply connected. The smoke is
simulated by tracking a large number of Lagrangian
elements, which originate in the fire. These same
elements carry the heat released by the fire, providing
a self consistent description of the smoke transport at
all resolvable length and time scales. Large
temperature and pressure variations are permitted,
subject to the limitation that the Mach number is much
less than one. The next two sections give a brief
description of the mathematical and computational
aspects of the model, while the final section
illustrates its capability with sample results and a
comparison with experiment.