Turbulent strømning og fordampning av væske i sjakt

The aim of this study is to contribute to a better understanding and fundamental knowledge on the behaviour of harmful chemical substances when released, and how they are dispersed in air. Such knowledge is a crucial foundation for estimating the consequences of possible chemical incidents, and for the planning of protective actions and countermeasures, and also for training and exercises for the Defence and for civilian emergency authorities. There are numerous models for calculating the dispersion and estimating hazard areas, and possible consequences of attacks, sabotage or accidents. A complete model includes a mathematical and physical description of a number of phenomena that constitute a dispersion incident, and can roughly be divided in source modelling, transport modelling, and effect modelling respectively. Improvements of such models is a field that still requires a lot of research, as there are large uncertainties in the three parts of the modelling chain. This study focuses on source modelling. FFI has contributed to the NATO study RTO-SAS-061 ’Defence against CBRN-attacks in the changing NATO strategic environment’ with a scenario of indoor dispersion of sarin. The scenario was also simulated by other participants, and it turned out that different participants used evaporation rates that were set apart with a factor of up to 10. This gives correspondingly large variations on the calculations of the consequences and the extent of damages, and is the background for this study. An accurate description of the evaporation is vital for a realistic modelling of the dispersion, both for time and concentration. The properties of the air flow above the evaporating liquid surface have a great influence on the evaporation rate. The turbulent boundary layer directly above the pool is particularly important. The object of this work is to study the evaporation from a liquid, in this case the nerve agent sarin, in a shaft, in order to contribute to a better description of the evaporation of liquids in general. This can be used both in combination with simple tools for estimating the hazard areas, simple dispersion models, and more complex models. In this work, the turbulent flow through a quadratic shaft has been simulated by means of various turbulence models in the Computational Fluid Dynamics code Fluent. The results are compared with direct numerical simulations in order to verify the chosen method. Based on the turbulent velocity fields from the simulations, the evaporation from a pool in the shaft and the subsequent dispersion of the vapour are simulated. This is especially important for indoor dispersion where ventilation systems are utilised, but also important in other scenarios that include evaporation of chemicals from a pool. The results show that the air flow perturbs the liquid surface, and this has a great effect on the evaporation rate. The evaporation rate calculated from the simulations is about four times greater than the corresponding evaporation rate calculated with a simple formula based on molecular diffusion through a laminar flow. They also show that some of the sarin binds to the walls of the shaft, and this causes a secondary evaporation. This work will be continued by examining other chemical substances and by varying the properties of the flow. The plan is to construct a data base with initial conditions for dispersion calculations, so that later calculations of evaporation can be executed quickly and accurate.