Sensitivity analysis and validation of a predictive procedure for high and low-temperature ignition delays under engine conditions for n-dodecane using a Rapid Compression-Expansion Machine

A predictive procedure for cool flames and high-temperature ignition delays based on the accumulation and consumption of chain carriers has been validated for n-dodecane under engine conditions. To do so, an experimental parametric study has been carried out in a Rapid Compression-Expansion Machine, measuring the ignition times for different compression ratios (14 and 19), initial temperatures (from 403 K to 463 K), O2 molar fractions (from 0.21 to 0.16) and equivalence ratios (from 0.4 to 0.7). The measured ignition delays have been compared to results from chemical kinetic simulations performed in CHEMKIN using a 0-D reactor that replicates the experimental conditions by solving five different chemical kinetic mechanisms, as a way to evaluate the mechanisms accuracy and variability. In general, all chemical kinetic mechanisms are able to accurately replicate the experimental ignition delays, being the mean relative deviation lower than 1.9% and 1.6% for both ignition stages, cool flames and the high-temperature ignition respectively. Furthermore, small differences have been appreciated between mechanisms in terms of ignition delay. Then, the predictive method has been applied using different databases obtained from each mechanism and a sensitivity analysis has been performed in order to evaluate the effects of the selected database on the predicted ignition delay. It has been found that while cool flames seems to be independent on the selected mechanism, the predicted high-temperature ignition delay is very sensitive to the species selected as chain carrier. Thus, if formaldehyde is assumed as ignition tracer, the predicted ignition time can vary up to 3%, while this percent decreases up to 1.3% when hydrogen peroxide takes the role of chain carrier.