A flexible and stable numerical method for simulating the thermal decomposition of wood particles

The objective of this paper is to present a flexible and stable simulation method to predict the thermal conversion of wood particles. A combination o f several subprocesses such as heating-up, drying, pyrolysis, gasification and combustion of fuel particles of different properties and sizes represen ts the global process of thermal conversion. This approach allows for simul taneous processes e.g. reactions in time and covers the entire range betwee n transport-limited (shrinking core) and kinetically limited (reacting core ) reaction regimes. Thus, the model is applicable to simulate sufficiently accurate the thermal decomposition of each particle in a packed bed, of whi ch the entire conversion is regarded as the sum of all particle processes. Effects such as fragmentation, swelling, homogeneous reactions e.g. ignitio n outside a particle are excluded as a tradeoff between complexity and calc ulation time. However, a description of the particle processes by one-dimen sional and transient differential conservation equations for mass and energ y is feasible to represent the above mentioned subprocesses. The particles are coupled to the gas phase by heat and mass transfer taking into account the Stefan correction due to the gas outflow during conversion. A general f ormulation of the conservation equations allows the geometry of a fuel part icle to be treated as a plate, cylinder or sphere. In order to achieve a hi gh degree of flexibility, the method distinguishes between data, such as ki netics or material properties and the conversion process, for which relevan t data are stored in a data base for easy access and extension. The resulti ng modules of this subdivision are encapsulated into separate software unit s cast in a hierarchy of well-defined classes in Tools of Object-oriented S oftware for Continuum-Mechanics Applications (TOSCA) by object-oriented tec hniques. (C) 2001 Elsevier Science Ltd. All rights reserved.