On the influence of initial conditions and facility effects on rapid compression machine data

Abstract

The influence of rapid compression machine (RCM) facility-effects on ignition delay time is investigated using multi-zone model simulations that consider the role of crevice volume and compression time. Four virtual machines are defined, each with a unique combination of crevice volume (∼5 to 7 mL) and compression time (25–50 ms). These machines are used to simulate autoignition of four fuels (iso-octane, n-heptane, PRF50, and ethanol) from low to high temperatures at equivalence ratios of ϕ = 0.5, 1.0, 2.0 and compressed pressures of pc = 10, 20 and 40 bar. This range of simulation conditions provides a diverse array of autoignition responses, including single-stage and two-stage autoignition. Furthermore, the simulation conditions are obtained for each machine by using various combinations of compression ratio, initial temperature, and argon:nitrogen fraction in the diluent gas. The results show that for a given machine, the test method used to obtain the compressed condition can strongly influence the ignition delay time, particularly under two-stage ignition conditions. The disparity between the test methods grows with the duration of the second stage of ignition, thus the test method dependence is most pronounced for iso-octane (i.e., PRF100). Comparison of the virtual machines shows that increased crevice volume (at roughly constant surface-area-to-volume ratio) leads to enhanced heat loss and therefore longer ignition delay times. For short ignition delay times (<10 ms), lengthening the compression time from 25 to 50 ms initiates reactivity prior to the end of the stroke, thus leading to a reduction in the ignition delay time. However, the simulations reveal that even for long ignition delay times, in excess of 100 ms, the ignition delay times are still sensitive to the compression time due to crevice heat loss dynamics that occur in the first ∼20 ms after the end of compression. A comprehensive comparison of the data suggest that variability among unique machines may be minimized by making measurements at low-to-moderate compression ratios and avoiding RCM designs with small crevice volumes.