Optimized rate rules and kinetic models for the oxidation of large normal and iso-alkanes

Accurate kinetic modelling of large aliphatic alkanes is essential for fossil and sustainable fuel combustion; however, a comprehensive set of reliable rate rules to construct predictive models is still lacking. In this work, a set of rate rules was optimized using a comprehensive data collection of ignition delay time (IDT) and species concentration measurements for 25 linear and branched C 5 −C 12 alkanes. By reviewing theoretical studies of alkane reactions, prior uncertainty ranges of Arrhenius-type rate coefficients for the rate rules were determined. Based on the central Arrhenius curves of these ranges, the initial mechanisms of these fuels were autogenerated using the MAMOX++ code and merged into an 8,395-species and 27,376-reaction model, building upon the C 0 –C 4 mechanism from C3MechV4.0.1 as the core mechanism. The thermodynamic data for larger alkane species (C >4 ) were taken from the high-level theoretical calculations of Elliott et al. or estimatedusing the group additivity method trained on the Elliott et al. data. As optimization targets, we collected 1,484 and 699 IDTs measured in shock tubes, rapid compression machines, respectively, and 2,473 species concentrations measured in jet-stirred reactors. To make the optimization feasible, we employed the Model-Reduction-Assisted Parameter-Optimization-Based Mechanism Development (RAPOD) approach using the DRGEP mechanism reduction method, and introduced novel modifications for multi-fuel model optimization. Using this fuel-wise RAPOD (f-RAPOD) method, accurate reduced mechanisms were developed for each of the 25 fuels, comprising 379–1,670 species and 2,637–7,172 reactions, which allowed simulations to run up to 21 × faster. The Arrhenius parameters associated with 255 rate rules were optimized within their prior uncertainty ranges using the Optima++ code. During optimization, the parameters were uniformly updated across all reduced models, and simulations of single-fuel experiments were carried out using the corresponding fuel-specific reduced model. A comparison of the optimized model with literature mechanisms shows that the present model achieves the lowest overall prediction error for almost all fuels and experimental types, indicating superior predictive performance. The optimized rate constants exhibit physically consistent trends and align well with expected chemical behaviour, confirming the reliability of the adjusted rate rules. Novelty and significance statement This work represents the first comprehensive optimization of rate rules for the oxidation and pyrolysis of large aliphatic alkanes, covering the full spectrum from linear to highly branched isomers. The optimized rate rules provide a reliable reference framework for constructing combustion kinetic models and can be feasibly extended to other fuel classes ( e.g. , biodiesels) by incorporating specific functional group chemistries. Consequently, this study lays a foundation for the further development of reaction mechanisms for cyclic, unsaturated, and oxygenated hydrocarbons. In addition, we propose robust, validated reaction mechanisms for 25 large aliphatic alkanes ranging from C 5 to C 12 . A novel fuel-wise Model-Reduction-Assisted Parameter-Optimization-Based Mechanism Development (f-RAPOD) method is proposed and applied in combination with other acceleration techniques and advanced computational tools to enable efficient optimization, demonstrating a scalable approach for future large-scale combustion kinetic optimization problems.