Computer modelling and associated problems

For the last 20 to 25 years, computer modelling has been used increasingly to interpret combustion phenomena. When the chemistry is the main interest, for example, in predicting ignition delay times or predicting product profiles, large comprehensive mechanisms can be used which in theory cover every species possible in the total oxidation. In most treatments differential equations are written for the reactants, intermediate species and for the final products. These are solved by standard mathematical treatments and profiles of concentration - time produced for all species. This approach will be discussed in detail in Chapter 4. 

An obvious problem is that a rate constant value is needed for each reaction in the mechanism in order that the reaction may be included in the computer calculation. Unfortunately there is a serious lack of kinetic data for elementary reactions, many of which have not been studied experimentally. This is particularly true of reactions associated with large alkyl radicals, which are of prime importance in the combustion of practical fuels. Three other pertinent difficulties arise. 

Frequently, kinetic data determined at low pressures require lengthy extrapolation to engine conditions (ca 20 atms or above). Fundamental theories are available but very difficult for modellers to use. Fortunately, user-friendly methods have been developed by Gilbert et al. [7] and Baulch et al. [8]. 

Although k = 6.0x 10 cm mol s is known accurately for the overall reaction between 300 and 2000 K, the relative importance of the channels is not clear, which severely restricts rigorous modelling. 

For an elementary reaction the activation energies of forward and reverse steps are linked to the enthalpy of reaction, so that Ef- Et, = AH. 

---