Methods and applications of pyrolysis modelling for polymeric materials [Pyrolyysimallinnuksen metodeita ja sovelluksia polymeereille]

Fire is a real threat for people and property. However, if the risks can be identified before the accident, the consequences can be remarkably limited. The requirement of fire safety is particularly important in places with large number of people and limited evacuation possibilities (e.g., ships and airplanes) and for places where the consequences of fire may spread wide outside of the fire location (e.g., nuclear power plants). The prerequisite for reliable fire safety assessment is to be able to predict the fire spread instead of prescribing it. For predicting the fire spread accurately, the pyrolysis reaction of the solid phase must be modelled. The pyrolysis is often modelled using the Arrhenius equation with three unknown parameters per each reaction. These parameters are not material, but model specific, and therefore they need to be estimated from the experimental small-scale data for each sample and model individually. The typical fuel materials in applications of fire safety engineers are not always well-defined or characterised. For instance, in electrical cables, the polymer blend may include large quantities of additives that change the fire performance of the polymer completely. Knowing the exact chemical compound is not necessary for an accurate model, but the thermal degradation and the release of combustible gases should be identified correctly. The literature study of this dissertation summarises the most important background information about pyrolysis modelling and the thermal degradation of the polymers needed for understanding the methods and results of this dissertation. The articles cover developing methods for pyrolysis modelling and testing them for various materials. The sensitivity of the model for the modelling choices is also addressed by testing several typical modeller choices. The heat release of unknown polymer blend is studied using Microscale Combustion Calorimetry (MCC), and two methods are developed for effectively using the MCC results in building an accurate reaction path. The process of pyrolysis modelling is presented and discussed. Lastly, the methods of cable modelling are applied to a large scale simulation of a cable tunnel of a Finnish nuclear power plant. The results show that the developed methods are practical, produce accurate fits for the experimental results, and can be used with different materials. Using these methods, the modeller is able to build an accurate reaction path even if the material is partly uncharacterised. The methods have already been applied to simulating real scale fire scenarios, and the validation work is continuing.