Reduction of large detailed chemical kinetic mechanisms for autoignition using joint analyses of reaction rates and sensitivities

Independent technical summary

The study presents a reduction method for detailed autoignition mechanisms that combines reaction-rate analysis with sensitivity analysis to identify reactions that can be removed while preserving selected target behaviour of the parent mechanism within the reduction and evaluation domain.

The work examines how threshold values used in the two analyses affect the resulting mechanism size, computational efficiency and predictive accuracy. It also develops practical rules for selecting those thresholds and applies the method to detailed mechanisms for two reference hydrocarbons: n-heptane and isooctane.

The reduced mechanisms were reported to reproduce the selected ignition-delay and intermediate-species results of the parent mechanisms with good agreement over the evaluated cases. The article reports calculation speedup factors of 5.9 for the n-heptane mechanism and 16.7 for the isooctane mechanism. These speedup values and fidelity statements apply to the study's defined mechanisms, observables, numerical formulation and test conditions and should not be treated as universal performance factors.

Technical contribution

The work addresses a central modelling problem in combustion kinetics: reducing the computational cost of detailed mechanisms while retaining enough chemical structure to reproduce selected autoignition observables within a defined application domain.

Its principal methodological contribution is the joint use of reaction-rate and sensitivity information rather than reliance on a single reduction criterion. The resulting reduced mechanism remains application-dependent because retained fidelity is defined against the parent mechanism, fuels, thermochemical conditions, reactor formulation and target quantities used during reduction and evaluation. Agreement with the parent mechanism is computational validation of the reduction relative to that parent model; it is not by itself experimental validation of either mechanism.

Scope and application boundary

A reduced mechanism should not be assumed to retain equal fidelity outside its evaluated fuel compositions, temperature and pressure ranges, equivalence ratios, reactor assumptions and target observables. A reduction optimized for ignition delay may not retain the species or pathways required for flame propagation, emissions, soot precursors or other objectives.

Application to CFD, engine simulations, new fuels or substantially different conditions requires re-evaluation against the intended operating domain. This may include ignition delay, heat release, flame behaviour, major and intermediate species and pollutant or soot precursor pathways relevant to the new use case, together with experimental validation where predictive physical use is claimed.

Evidence interpretation

Three evidence levels should be kept distinct. First, the reduction algorithm can be verified as an implementation. Second, a reduced mechanism can be compared computationally with its parent detailed mechanism for selected targets. Third, either mechanism can be compared with independent experiments. Success at one level does not automatically establish the next.

Mechanism size reduction and calculation speedup are therefore useful engineering outcomes only when the retained chemistry remains adequate for the intended decision. The smallest mechanism is not necessarily the most useful mechanism.

Publisher access

The Version of Record is available from Wiley through the DOI: access the publisher record . Access may require an institutional or individual subscription.

Citation

Saylam, A., Ribaucour, M., Pitz, W. J., & Minetti, R. (2007). Reduction of large detailed chemical kinetic mechanisms for autoignition using joint analyses of reaction rates and sensitivities. International Journal of Chemical Kinetics, 39(4), 181–196. https://doi.org/10.1002/kin.20232

Relation to later kinetic-intelligence work

This 2007 study is a specific mechanism-reduction method based on joint reaction-rate and sensitivity analyses. Later work on reaction networks, degree centrality, DRG, adaptive chemistry and state transfer addresses related but distinct reduction and analysis problems. Those methods should not be presented as if they were part of, or validated by, this original study unless an explicit comparison establishes that connection.

Related technical resources

• Modelling Studies of the Oxidation and Autoignition of Alkanes, Aromatics, and Their Mixtures at High Pressure
• HCCI Engine as Chemical Reactor to Produce Fuel/Chemicals
• Methodology to Reduce Diesel Engine Pollutant Emissions
• Kinetic Intelligence — reduction, network analysis and adaptive chemistry