Simultaneous side-wall-schlieren and -emission imaging of autoignition phenomena in conventional and constrained-reaction-volume shock-tube experiments

Abstract

Simultaneous schlieren and emission imaging through the side wall of a round shock tube is reported for application to experimental autoignition studies. A round, optically accessible shock tube featuring windows designed as aberration-corrected cylindrical lenses provides a large side-wall field of view compatible with both emission and schlieren imaging. Autoignition experiments are reported for non-dilute propane–oxygen–argon mixtures (ϕ=1 C3H8 in 21% O2, 79% Ar) at temperatures (T5) spanning the range 1,203–1,438 K and pressures near 1 atm. Experiments are performed using both conventional and constrained-reaction-volume (CRV) filling, allowing for the most detailed comparison to date of ignition dynamics between these two types of experiments. When T5 exceeds 1,290 K, strong ignition is observed to initiate in the immediate proximity of the end wall regardless of the filling strategy. In lower-T5 experiments, mild ignition is observed to initiate spontaneously at remote locations, which often exceeding the imaging FOV in conventional experiments but are constrained within about 5 cm of the end wall in CRV experiments. Measured ignition delay times (IDTs) are compared to predictions from three kinetic mechanisms. IDTs simulated using the San Diego and USC Mech II mechanisms are of the same magnitude as the measurements but exhibit a lower activation energy; the propane submechanism from NUIG 1.1, meanwhile, systematically over predicts observed IDTs by roughly a factor of two. Comparisons are also made between the experimental measurements and predictions from a recently introduced 1-D axial-temperature-gradient model describing remote ignition in shock tubes. Considering the effect of a supposed dT5/dz=150 K/m gradient, the 1-D model functionally predicts the remote ignition observed in T5< 1,290 K experiments and suggests the use of CRV filling may significantly reduce the error introduced into IDT measurements by remote ignition at low-T5 conditions.