Computer Fire Models
by Anthony D. Putorti, Jr.
Fire Safety Engineering Division
Building and Fire Research Laboratory
National Institute of Standards and Technology
Computer fire models are compilations of mathematical
equations derived from basic physical principles or experimental data. The
computer is used to solve systems of equations that would be difficult or
tedious to solve by hand. The simplest computer models solve single equations,
while more complex models are composed of hundreds or thousands of equations.
A specific type of fire model, the room fire model, is capable of predicting
the development of fire conditions in structures and is a helpful tool for
fire investigators. Fire models are useful for predicting the development
of a fire in a room or structure, for evaluating fire scenarios developed
by an investigator, for comparing fire events to established time lines,
and for conducting what-if analyses.
Due to the wide range of fire models available,
varying levels of expertise are necessary to properly apply the models to
investigations. Users of fire models must have an in-depth knowledge of
the specific assumptions made by the models and the origins of the experimental
correlations and data used as inputs. Each model will have specific limitations
as a result of the assumptions made and the experimental methods used to
derive correlations and input data.
There are two major categories of room fire models
available to fire investigators: zone models and field models.
Zone models split rooms or enclosures up into one
or more zones. The most commonly used models assume a room is made up of
two zones; an upper layer consisting of heated combustion products, and
a lower layer which is composed of cooler air relatively free of combustion
products. In a two zone model, the fire forms the connection between the
upper and lower layers. The layers are assumed to be well mixed, so that
the conditions within each layer are constant. The predicted temperature
within the hot layer, for example, is the same throughout. Many of the models
include provisions for openings to the outside or to other rooms and for
heat losses to the walls and ceiling. Model inputs typically include room
dimensions and building materials, the sizes and locations of room openings,
room furnishings characteristics, and the fire heat release rate. Outputs
typically include prediction of sprinkler or fire alarm activation time,
time to flashover, upper and lower layer temperatures, the height of the
interface between the upper and lower layers, and combustion gas concentrations.
Zone room fire models are available from several sources including the National
Institute of Standards and Technology (NIST).
Field models, also known as computational fluid
dynamics models (CFD), split a room or enclosure up into a large number
of small three dimensional boxes called cells. The enclosure may contain
hundreds of thousands of cells ranging in size from centimeters to meters.
Field models are based on the basic physical principles of energy, mass,
and momentum conservation. The computer calculates the movement of heat
and smoke between the cells over time. At any point in time, it is possible
to find the temperature, velocity, and gas concentrations within each of
the cells. As in the case of the zone models, the properties within each
cell are assumed to be constant. Due to the larger number of cells however,
the conditions in the enclosure can be predicted in much greater detail.
Field models are capable of predicting the conditions in very large and
very small spaces, in spaces with complex shapes, and in complex multiple
room configurations that are not possible with zone models. Due to the complexity
of field models, they require a high level of expertise to operate and are
currently run on expensive computer equipment. General purpose field models
are available commercially from various sources. Advanced field models,
specifically designed for fire safety analysis, are under development at
the National Institute of Standards and Technology (NIST).
Reprinted with permission from the author.