On the application of planar laser Rayleigh scattering imaging to Ethylene-air detonations at normal atmospheric conditions

Detonation-based technologies have been developed primarily by evaluating different design concepts using global performance metrics (e.g., wave speeds and thrust output). However, several technical challenges, which may require a physics-based development approach to mitigate or resolve, continue to hinder these systems. Unfortunately, the most commonly used tools for characterizing the physics of detonations (i.e., cell size measurements, empirical correlations, and predictive models) don’t typically provide enough detailed information to make a significant impact on the development of detonation-based technologies. Therefore, this study demonstrates the use of planar laser Rayleigh scattering (PLRS) imaging to acquire highly resolved data from ethylene-air detonations produced at atmospheric conditions. In particular, the PLRS images enable rare measurements of induction zone regions and allow the details of a detonation’s underlying structure to be visualized in great detail. The accuracy of the induction zone length measurements is assessed via a systematic numerical analysis involving the simulation of laser Rayleigh scattering (LRS) signals from one-dimensional (1-D) ZND results. The numerical results were quantitatively compared to measured LRS signal levels, their spatial gradients, and induction zone lengths at local wave speeds estimated from high-speed (MHz) chemiluminescence imaging. Varying levels of agreement between the measured and calculated results are observed, with the best agreement achieved for detonations propagating at or faster than the CJ speed. Beyond the measured results, this manuscript clearly highlights the benefits, challenges, and future potential of applying PLRS imaging to detonations. • Results from the 1-D ZND calculations indicated that computed induction length ( Δ i ) values can vary by up to two orders of magnitude depending on the chosen chemical mechanism. Such large discrepancies are most prevalent at equivalence ratios of ϕ ≤ 0 . 7 and local wave speeds of D / D C J ≤ 0 . 9 . • Simulated Rayleigh scattering signals, derived from the thermo-chemical state of C 2 H 4 - air detonations at NTP, clearly indicate that leading shock fronts are followed by regions with elevated Rayleigh signal levels. The lengths of these regions are within 6% of the Δ i values computed by ZND when ϕ ≲ 1 . 0 and D / D C J ≲ 1 . 05 . Although large errors accompany PLRS-based Δ i values when D / D C J > 1 . 1 , this work demonstrated that PLRS imaging can provide accurate induction zone length measurements within the other portions of the detonation cell cycle. Furthermore, the threshold level that defines Δ i can be tuned to sacrifice some measurement accuracy when D / D C J ≲ 1 . 05 for improved accuracy when D / D C J > 1.1. • An advantage of PLRS imaging is that its high spatial resolution can capture intricate details of detonation front structures. These included clear visualizations of nascent Mach stems after triple-point collisions and shocked, unburned pockets of reactants. Even at C- J conditions in detonations without substantial velocity deficits, shocked reactants were observed at distances behind the shock front of more than 10 × Δ i . • Quantitative comparisons between measured and simulated profiles of Rayleigh scattering signals are within reasonable agreement given limitations in the accuracy of the local wave-speed measurements and uncertainty associated with the chemical mechanisms. However, the measured results consistently produced lower Rayleigh scattering signal levels than anticipated from the calculations. • Values of Δ i derived from the PLRS images ( Δ i ,R) were found to vary substantially. This is consistent with the highly irregular nature of the detonations considered in this study. The measured values of Δ i ,R tended to exhibit better agreement with those derived from ZND calculations in regions away from triple points where D / D C J ≥ 1 . 0 . Yet, near triple points or when D / D C J < 1 . 0 , measured values of Δ i ,R did not agree well with those predicted by the ZND calculations.