Probing the Reaction Zone of Nanolaminates at ∼μs Time and ∼μm Spatial Resolution

Abstract

Reactive nanolaminates are a high-energy-density configuration for energetics that have been widely studied for their tunable energy release rates. In this study, we characterized Al/CuO nanolaminate reactions with different fuel/oxidizer ratios and bilayer thicknesses using both macro- and microscale high-speed imaging/pyrometry. Under microscopic imaging, we observe significant corrugation (the ratio of the total geometrical length of the flame to the width of the sample in the direction perpendicular to propagation) of the flame, which can increase the reaction surface area by as much as a factor of 3. This in turn manifests itself as an increase in the global burn rate (total nanolaminate film length/total burn time). We find that the global burn rate can be predicted as the product of the microburn rate (local vector burn rate at the microscopic scale) and the corrugation. These corrugation effects primarily impact fuel-rich conditions, resulting in higher global burn rates. We find that the reaction zone has a thickness of ∼150 μm. Finally, we present a 3D rendering of what we believe the reaction zone looks like, based on the results from in-operando observation and SEM cross-sectional imaging.

How Thermal Aging Affects Ignition and Combustion Properties of Reactive Al/CuO Nanolaminates: A Joint Theoretical/Experimental Study

“…Table 3 assembles a list of unpublished and recently published experiments [ 31 , 33 ] performed on sputter-deposited Al/CuO nanolaminates produced by our team and characterized in combustion. For both experimental and theoretical data, we have considered fresh nanolaminates, i.e., characterized just after sputter deposition, and aged nanolaminates, i.e., being annealed at 200 °C for 14 days.…”

“…To explain this, one has to note that the macroscopic burn rate corresponds to the global measurement of the reaction front velocity in the Al/CuO system. As discussed in [ 33 ], it corresponds to measuring the microscopic (local) burn rates at the reaction front multiplied by the corrugation of the flame front, which is the ratio of the total geometrical length of the flame to its width. While the exact origin of corrugation is still elusive, it is clearly the result of mechanisms different from the pure condensed phase combustion that fundamentally operates in our propagation model.…”

“…The significant corrugation for fuel rich cases leads to a larger burning surface area (× ~3) and, consequently, faster observed burn rate (macroscopic) in fuel rich samples (details in [ 33 ]). Hence, considering microscopic burn rate, measured by microscale high speed imaging (not shown in this article but detailed in [ 33 ]) and with respect to fresh samples, fair agreement is found with the experimental values, being of 4 and 3.5 m·s −1 for ξ 2 and ξ 3 stacks. For aged samples, comparison of fuel rich samples is not straightforward, as the corrugation and microscopic burn rate were not characterized.…”

Modeling the self-propagation reaction in heterogeneous and dense media: Application to Al/CuO thermite

“…Other performed DFT (Density Functional Theory)-based MD to model the onset of Al/Fe2O3 and Al/CuO thermites [46,47]. While MD simulations are a valuable means of depicting atomic-scale processes, the duration of the process that can be modeled remains short, from pico-to nanoseconds, which definitively limits their application to describe the flame front combustion dynamics over several µm thick [48]. Other attempts were proposed to extend a simple continuum approach developed for bimetallic reactive materials [49] based on the sandwich theory where overall mass transport and the reaction is assumed to follow a single Arrhenius dependence on temperature, independently of the detailed chemical composition.…”

“…But considering a single mobile species, namely oxygen, diffusing across the various system layers, does not fit the complexity of the mechanisms occurring in thermite reactions, notably in the high temperature regime. Moreover, the 1D description of the flame propagation is a severe limitation to simulating the reaction front kinetics, for instance in treating inhomogeneities (additives, porosity), making this modeling approach unsuitable for capturing many of the existing experimental data in the field [48]. Hence, we propose a new 2D nonstationary model, based on CFD schemes and implementing more realistic reaction mechanisms that consider both oxygen and Al diffusion, since it was experimentally evidenced [52] but never explicitly introduced in thermite modelling to date.…”