Physico-chemical based models for the prediction of safety relevant ignition processes

Detailed numerical simulations of combustion processes using the complete set of time-dependent governing equations and complex chemistry have become an important research tool in the recent years. Whereas in most cases, such simulations are used to optimize practical combustion systems like internal combustion engines or gas turbines. These simulations are also of tremendous value to explain the multiple-parameter dependence of ignition hazards. Turning such simulations from very generic cases with relatively simple geometries to more complex technical systems, can lead to a paradigm shift from empirical knowledge based on safety characteristic data towards the prediction of safety-related ignition processes.

In the talk, we will discuss the principle of these modelling strategies and their application to generic ignition scenarios like auto ignition, ignition by sparks or by hot particles. Although these ”generic” scenarios already allow a good insight into the governing processes, it is important for practical applications to characterize the overall ignition process. Therefore, we shall discuss in the second part of this work how the detailed information can be used to devise models for the overall ignition process. The problem in modelling these (typically turbulent) processes is that the description of chemically reacting systems leads to scaling problems in space and time. In particular, an oversimplification of the coupling processes between chemical reaction and turbulent flow should be avoided by all means to allow a predictive character. In the presentation, it is shown how hierarchical concepts can be used to solve this problem.