Autoignition and preliminary heat release of gasoline surrogates and their blends with ethanol at engine-relevant conditions: Experiments and comprehensive kinetic modeling

This work utilizes a rapid compression machine (RCM) to experimentally quantify autoignition and preliminary heat release characteristics for blends of 0 to 30% ethanol by volume into two surrogates (FGF-LLNL and FGF-KAUST) that represent a full boiling range gasoline (FACE-F). Experimental conditions cover pressures from 15 to 100 bar, temperatures from 700 to 1000 K, and diluted/stoichiometric and undiluted/lean fuel loading conditions representative of boosted spark-ignition and advanced compression ignition engines, respectively. Direct comparison is made with previously reported results for FACE-F/E0–E30 blends. A detailed gasoline surrogate chemistry model is also proposed, and chemical kinetic modeling is undertaken using the proposed model to generate chemical insights into the compositional effects and ethanol blending effects. Although experiments show similar qualitative trends between the surrogates, quantitative differences between the surrogates are obvious, where FGF-LLNL displays greater low-temperature reactivity and faster evolution of low-temperature heat release (LTHR) than FGF-KAUST, with such differences being significantly muted by ethanol blending. Flux analyses reveal the compositional effects on surrogate reactivity at the diluted/stoichiometric condition, where n-heptane facilitates first-stage ignition reactivity for FGF-LLNL/E0 by initiating earlier and more rapid ȮH branching than n-butane for FGF-KAUST/E0. Sensitivity analyses highlight the importance of non-fuel-specific interactions between ethanol and surrogate sub-chemistries in controlling the reactivity of ethanol-blended surrogates. Direct experimental comparisons between the surrogates and FACE-F, as well as between the surrogate/EtOH and FACE-F/EtOH blends highlight the need of high-fidelity surrogates that can fully replicate the target gasoline in properties including ignition reactivity and LTHR characteristics at extended conditions, as well as their response to ethanol blending. Overall, the model captures the experiments reasonably well. Nevertheless, the model displays increasing disagreement with experiments for the two surrogates at higher levels of ethanol blending, and this is found to be caused primarily by non-fuel-specific interactions between ethanol and surrogate component sub-chemistries. Futhermore, the inadequacy of the kinetic model to capture surrogate-to-surrogate differences at the diluted/stoichiometric condition suggests more physical testing is needed to facilitate more extensive model validation.