Determination of the material parameters is one of the key challenges of numerical fire simulation attempting to predict, rather than prescribe the heat release rate. In this work, we use common fire simulation software and genetic algorithms to estimate the kinetic reaction parameters for wood components, birch wood, PVC and black PMMA. Parameters are estimated by modelling thermogravimetric experiments and minimizing the error between the experimental and numerical results. The implementation and choice of the parameters for the genetic algorithm as well as the scheme to describe wood pyrolysis are discussed.
A new experimental apparatus for measuring flame spread rates at different ambient temperatures is presented. The 2 m long sample is pre-heated with air to the desired temperature, ignited from below with a small propane burner, and flame spread is monitored with thermocouples at the surface of the sample. The rate of vertical concurrent flame spread as a function of ambient temperature is determined on cylindrical Birch rod samples and on polyvinylchloride cable samples. Corresponding flame spread scenarios are numerically simulated using an axi-symmetric solution of the flow field and a pyrolysis model with parameters estimated from thermogravimetric and cone calorimeter experiments. The simulation model was able to predict the flame spread rates within the uncertainties associated with the experiments and postsimulation analysis of the spread rate.
Three dimensional simulations of forest fire and wind induced flows are very time consuming because the size of the computational problem becomes very large. From the economical viewpoint, the use of twodimensional simulations is an attractive alternative as means to reduce the computational cost by one or two orders of magnitude. However, the real fires and turbulent flows are never really two-dimensional, and making such simplification may introduce errors whose magnitude is not well known in advance. In this work, the effect of the 2D assumption is studied by performing a series of simulations in both two and three dimensions, and reporting the difference in the effects on structures. The results show that 2D simulations can be used for order-of-magnitude type of analysis, for which purpose they are well suited due to the small computing times. However, the differences seem to be too large for accurate predictions of the building and human response. Therefore, the critical simulations of the future analysis should be made in three dimensions.
KEYWORDS
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.