Reference
Fire Modeling - Fashionably Hot Analyses
By Mark Hughes, P. Eng.
© 2003 Sintra Engineering Inc.
Fire modeling is an area of fire spread/development analysis that many fire investigators are not particularly familiar with, but it can be, on occasion, a useful tool particularly in complex litigation files. For those not familiar with fire models, the obvious question becomes "What is a fire model and what are the benefits of using one?"
A fire model is a representation of a fire scenario. Fire models can be divided into two main categories: physical models and computer models. Physical models are scale models of the fire used to evaluate certain properties of how fires and smoke movement develops, generally in compartments. Computer models simulate the same conditions within the confines of the virtual world within the computer. Computer models are increasingly gaining favour as a method of evaluating and analyzing larger scale fires. Although there is a tendency to think that technology will provide us with all the answers we seek, in truth sometimes simple scale (physical) models can provide more information than existing computer models.
Computer models are generally divided into three main types. There are zone models, computational fluid dynamics (CFD) models and application specific models. With the advent of more powerful computer systems, many of these fire models are easier to use and apply. What in the past required highly technical mainframe computers, now can be accomplished with the use of a powerful desktop PC. CFD models, in particular, require massive computer resources to simulate a fire.
Zone models work by dividing the compartment being examined into different zones. The computer model then evaluates how these zones interact with each other and how the fuel loads within the compartment interact with the space. These models can involve multiple compartments and many fuel packages within the space. There are limitations to zone models in terms of how the models handle the boundaries (walls and ceilings) of the compartments and how they interact with the fire.
CFD models are far more complex and use sophisticated algorithms to model every particle within a space. Just as with zone models, it is important to understand the heat release rates of the burning objects within the compartment in order to be able to properly model the fire and fire growth. Because of the complexity of CFD models, these types of models are difficult to stage and simulate.
Application specific models try to model very specific phenomena and have limited applications for anything other than their intended model. An example is a wall model, which models the heat transfer and growth through a specific type of wall. These models are generally either used for academic purposes to prove the results of a lab test or in conjunction with other types of models to simulate boundary conditions.
All computer models generally started out as academic studies to try and predict fire growth particularly in comparison to small simulated fires. All computer models require a fundamental understanding of heat release rates of the objects burning within the space. These heat release rates are specific to the objects that are burning and the rates heavily affect the results of all types of computer models. Significant errors in the assessment of the heat release rate for a burning object can produce results that diverge significantly from the actual fire.
All this technical information is really interesting (ho hum!) but in the end the question is what exactly are these models used for and how can they benefit the investigator? Well for most investigators, these models a still too complex and theoretical to provide any real benefit, but probably within the next few years as the models get more sophisticated and the computers get faster, their general use will become more widespread. Models are currently used to analyze sprinkler activation, fire spread, evacuation paths, smoke development and fire growth. These are useful to builders, architects and fire departments. As these models get easier to use, they will eventually allow the average investigator to test their theories for how a fire started.
For now this is still the stuff of Star Trek and requires a lot of time
and understanding of the physics of fire dynamics, but maybe soon fire investigators
will be able to use these tools to "boldly go where no fire investigator
has gone before". Sorry for the bad joke.
return to Reference articles...
