On the facility independence of inhomogeneous ignition processes in shock tubes: propane/air ignition at low temperature and intermediate pressure

Abstract

The interpretation of fuel/air ignition delay times measured in shock-tube experiments becomes challenging at low temperatures and low dilution levels due to inhomogeneous ignition phenomena across the test gas. The rapid growth of a turbulent boundary layer at the wall of the shock tube introduces non-ideal, gas-dynamic effects that can disturb the conditions behind the reflected shock wave, especially at longer test times, potentially causing localized ignition. The magnitude of such effects may vary between shock-tube facilities, but their influence on ignition may not necessarily be facility-dependent. To better understand this phenomenon, ignition delay time and high-speed chemiluminescence imaging measurements for lean and stoichiometric C₃H₈/O₂/diluent mixtures have been performed in two different shock-tube facilities at two different laboratories that differ in geometry. C₃H₈/O₂ mixtures were diluted in both N₂ and Ar with diluent/O₂ volume ratios maintained at 3.76/1. Temperatures behind the reflected shock wave ranged from 900 to 1400 K at reflected-shock pressures of 6 to 8 bar. Mixture-averaged dp/dt values are reported at 0.5 to 2.0 %/ms in the N₂-diluted mixtures and up to 4.0 %/ms in the Ar-diluted mixtures. The ignition delay time results are compared to simulations based on the NUIGMech 1.3 mechanism under both constant-volume conditions and incorporating facility-dependent dp/dt. Measured ignition delay times begin to depart from model predictions at 1050 K, forming a ‘roll-off’ region, where inhomogeneities accelerate ignition. These non-uniformities were first observed in both facilities at UDE and TAMU using pressure and chemiluminescence diagnostics and later confirmed at TAMU with endwall high-speed imaging during ignition. The measurements reveal that flame kernels appear up to 3 ms before the main ignition event, ultimately leading to shorter ignition delay times when compared to model predictions. Additionally, localized ignition appears to occur at the sidewall first, where larger temperature non-uniformities may exist due to interactions between the shock-heated gas and the turbulent boundary layer. This work extends inhomogeneous ignition studies in shock tubes by showing that such nonideal behavior is reproducible at similar conditions in two different shock-tube facilities from two different laboratories. Our results indicate that fluid-mechanic instabilities considerably influence the ignition process for post-reflected-shock temperatures below 1225–1326 K, depending on the criteria imposed on the definition of a bifurcation timescale. Experiments conducted in Ar-diluted mixtures containing 10 and 20 percent helium by volume were generally inconclusive, and it was determined that greater quantities of helium may be necessary to promote homogeneous ignition at the conditions of interest.