Development of a generalised kinetic model for the combustion of hydrocarbon fuels
Master Thesis
2010
Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
University of Cape Town
Department
License
Series
Abstract
The aim of this work is to find a generalised model for the combustion of hydrocarbons. Predicted temperature-time profiles can be obtained from detailed combustion kinetics, which can be used to derive a generalised model. If the generalised model can predict results from the detailed model it can be applied in computational fluid dynamics code where detailed kinetic mechanisms cannot.A generalised kinetic model is proposed, adapting the Schreiber model (Schreiber et al., 1994) to accurately predict the combustion behaviour of hydrocarbon fuels. The combustion behaviour is described through the characteristics of the temperature-time profiles and the ignition delay diagram, which include two stage ignition and the negative temperature co-efficient region. The Schreiber model is specifically adapted to improve the description of the very low temperature rise before and between ignitions and the auto-catalytic temperature rises during ignition. Using a Genetic Algorithm to optimise the prediction of the proposed model, the pre-exponent factor Ai and the activation energy Eai are the adjustable parameters which are optimised for each reaction in the model. These parameters have been optimised for three fuels: i-octane, n-heptane and methanol. The ignition delays of the pure fuels were accurately predicted. The temperature-time profiles in the instances of two stage ignition are relatively inaccurate. The temperature profiles are however an improvement on the temperature profiles predicted by the Schreiber model, particularly in terms of the slow temperature rise during the ignition delay andthe sharp temperature rise during ignition. The combustion of the binary blends of the three fuels have been predicted using model parameters which are found using the rate constants of each fuel, the blends composition and binary interaction rules. The binary interaction parameters were also optimised using a Genetic Algorithm. The binary interaction rules are based on the Peng-Robinson mixing rules. Overall the ignition delays of binary fuel blends were accurately predicted using binary interactions. However, when modelling the blends between methanol and n-heptane, where one fuel has extreme NTC behaviour and the other fuel has no NTC behaviour, the predictions were less accurate. These binary interaction rules are then used to model ternary mixtures. It is shown that the combustion behaviour of ternary mixtures of the three fuels can be accurately predicted without any further regression or parameter fitting. The accuracy of the ternary prediction is dependent on the accuracy of the binary predictions.
Description
Includes abstract.
Includes bibliographical references (leaves 73-76).
Keywords
Reference:
Fry, M. 2010. Development of a generalised kinetic model for the combustion of hydrocarbon fuels. University of Cape Town.