|
The
International Journal of Ventilation
Volume 3 No 2 September 2004:
Paper 7: Comparison of a CFD Fire Model against a Ventilated Fire
Experiment in an Enclosure
Yunlong Liu[1],
Alfred Moser[2] and Yehuda Sinai[3]
[1] CSIRO
Fire Science and Technology Laboratory, PO Box 310 North Ryde, NSW 1670,
Australia
[2] Air
and Climate Group, Department of Architecture ETH-Zentrum,
WET A5, CH-8092 Zurich,
Swiss Federal Institute of Technology, Switzerland
[3] CFX
Ltd., ANSYS CFX, The Gemini Building, Fermi Avenue, Harwell International
Business Centre, OX11 0QR, United Kingdom
Abstract
Computer modelling of fire has become an attractive approach for the fire
safety assessment of proposed building structures. To devise and validate the
fire model, a general-purpose computational fluid dynamics (CFD) software
package, CFX, has been evaluated against a fire test case in a ventilated room.
The test room is 6.0 m long, 4.0 m wide and 4.5 m high with an exit opening of
0.65 m x 0.65 m. A simple inert fire model is used, in which a constant
volumetric heat release is introduced at the location of the fire source to
represent the fire. Combustion chemical reactions are not included in the
computation; there was no flame spread and flashover to the wall lining material
in this experiment, as the wall lining was non-combustible. The heat
contribution is solely from the burner. It was demonstrated that both the k–
model and the Shear Stress Transport (SST) hybrid turbulence model are capable
of predicting the fire-generated turbulent flow and heat transfer. The
computational result has an error of about 20 °C when compared with the
measured gas temperature of the Lawrence Livermore National Laboratory (LLNL),
in which the heat release from a gas burner is used to represent a fire. It has
been confirmed that the thermal energy deposit into the wall plays a significant
role in the whole transient process. From the numerical point of view, the
exterior wall thermal boundary condition treatment has little influence on the
fire inside the room, as at 20 minutes after the start of the fire, heat
penetrates only about 3 – 4 centimetres into the 20-cm-thick wall in this
experimental case. The energy budget has been analysed to understand the energy
transfer for this fire test case. According to this test, about 30 percent of
the heat from the fire is released by thermal radiation, and about 30 percent is
carried out of the room by the ventilation air over the first 20 minutes after ignition, the rest
is deposited into the walls,
ceiling, and the floor.
It can be concluded that this CFD approach can serve as a tool for the modelling
of fire generated heat transfer in an enclosure. Thermal radiation plays an
important role in the heat transfer process from the fire. It has been concluded
that, in order to accurately simulate a fire case, the conjugate heat transfer
must be included in the fire mathematical model.
Key words:
Fire, CFD model, transient,
thermal radiation, conjugate heat transfer, turbulence, validation.
|
|
