The calculation of the consequences of fire occurring in a tunnel requ
ires a predictive model that can calculate smoke concentrations, tempe
ratures and radiation heat fluxes. In this paper, results from a field
model developed at the University of Sydney are compared with data ob
tained from experiments on pool fires in a ventilated tunnel. The mode
l solves conservation equations for mass, momentum and energy, togethe
r with those for the k-epsilon turbulence model. Combustion is modelle
d assuming that the chemistry is fast compared with the mixing, by sol
ving equations for the mixture fraction and its variance. Radiant hear
exchange between the gas, soot and walls is modelled using the discre
te transfer method. The calculations have highlighted a number of inte
resting features of the flow behaviour and modelling sensitivities. In
particular, a study of the effect of the turbulence modelling and soo
t radiation modelling on the predicted temperature stratification and
smoke backflow has been performed Comparison of the calculated tempera
ture fields and flame shapes with the experimental data has shown good
agreement and has quantified the sensitivity of the predicted results
to these important modelling parameters. This study highlights the ut
ility of field modelling for the analysis of fives and smoke movement
in tunnels.