Burning Behaviors of Solid Propellants using Graphene-based Micro-structures: Experiments and Simulations

2018-12-21T19:31:13Z (GMT) by Shourya Jain
Enhancing the burn rates of solid propellants and energetics is a crucial step towards improving the performance of several solid propellant based micro-propulsion systems. In addition to increasing thrust, high burn rates also help simplify the propellant grain geometry and increase the volumetric loading of the rocket motor, which in turn reduces the overall size and weight. Thus, in this work, burn rate enhancement of solid propellants when coupled to highly conductive graphene-based micro-structures was studied using both experiments and molecular dynamic (MD) simulations.

The experiments were performed using three different types of graphene-structures i.e. graphite sheet (GS), graphene nano-pellets (GNPs) and graphene foam (GF), with nitrocellulose (NC) as the solid propellant.

For the NC-GS samples, propellant layers ranging from 25 µm to 170 µm were deposited on the top of a 20 µm thick graphite sheet. Self-propagating combustion waves were observed, with burn rate enhancements up to 3.3 times the bulk NC burn rate (0.7 cm/s). The burn rates were measured as a function of the ratio of fuel to graphite layer thickness and an optimum thickness ratio was found corresponding to the maximum enhancement. Moreover, the ratio of fuel to graphite layer thickness was also found to affect the period and amplitude of the combustion wave oscillations. Thus, to identify the important non-dimensional parameters that govern the burn rate enhancement and the oscillatory nature of the combustion waves, a numerical model using 1-D energy conservation equations along with simple fi rst-order Arrhenius kinetics was also developed.

For the GNP-doped NC lms, propellant layers, 500  30 µm thick, were deposited on the top of a thermally insulating glass slide with the doping concentrations of GNPs being varied from 1-5% by mass. An optimum doping concentration of 3% was obtained for which the burn rate enhancement was 2.7 times. In addition, the effective thermal conductivities of GNP-doped NC lms were also measured experimentally using a steady state, controlled, heat flux method and a linear increase in the thermal conductivity value as a function of the doping concentration was obtained.

The third type of graphene structure used was the GF - synthesized using a chemical vapor deposition (CVD) technique. The effects of both the fuel loading ratio and GF density were studied. Similar to the GNPs, there existed an optimum fuel loading ratio that maximized the burn rates. However, as a function of the GF density, a monotonic decreasing trend in the burn rate was obtained. Overall, burn rate enhancement up to 7.6 times was observed, which was attributed to the GF's unique thermal properties resulting from its 3D interconnected network, high thermal conductivity, low thermal boundary resistance and low thermal mass. Moreover, the thermal conductivity of GF strut walls as a function of the GF density was also measured experimentally.

Then as a next step, the GF structures were functionalized with a transition metal oxide (MnO2). The use of GF-supported catalyst combined the physical e ect of enhanced thermal transport due to the GF structure with the chemical effect of increased chemical reactivity (decomposition) due to the MnO2 catalyst, and thus, resulted in even further burn rate enhancements (up to 9 times). The burn rates as a function of both the NC-GF and MnO2-NC loadings were studied. An optimum MnO2-NC loading corresponding to the maximum burn rate was obtained for each NC-GF loading. In addition, thermogravimetric (TG) and differential scanning
calorimetry (DSC) analysis were also conducted to determine the effect of NC-GF and MnO2-NC loadings on the activation energy (E) and peak thermal decomposition (PTD) temperatures of the propellant NC.

In addition to the experimental work, molecular dynamics simulations were also conducted to investigate the thermal transport and the reactivity of these coupled solidpropellant/graphene-structures. A solid monopropellant, Pentaerythritol Tetranitrate (PETN), when coupled to highly conductive multi-walled carbon nanotubes (MWCNTs) was considered. The thickness of the PETN layer and the diameter of the MWCNTs were varied to determine the effect of PETN-MWCNT loading on the burn rates obtained. Burn rate enhancement up to 3 times was observed and an optimal PETN-MWCNT loading of 45% was obtained. The enhancement was attributed to the faster heat conduction in CNTs and to the layering of PETN molecules around the MWCNTs surface. Moreover, the CNTs remained unburned after the combustion process, con rming that these graphene-structures do not take part in the chemical reactions but act only as thermal conduits, transferring heat from the burned to the unburned portions of the fuel.

A long-pursued goal, which is also a grand challenge, in nanoscience and nanotechnology is to create nanoscale devices, machines and motors that can do useful work. However, loyal to the scaling law, combustion would be impossible at nanoscale because the heat loss would profoundly dominate the chemical reactions. Thus, in addition to the solid propellant work, a preliminary study was also conducted to understand as how does the heat transfer and combustion couple together at nano-scales.

First, an experimental study was performed to understand the feasibility of combustion at nano-scales for which a nano-scale combustion device called "nanobubbles" was designed. These nanobubbles were produced from short-time (< 2000 µs) water electrolysis by applying high-frequency alternating sign square voltage pulses (1-500 kHz), which resulted in H2 and O2 gas production above the same electrode. Moreover, a 10 nm thick Pt thermal sensor (based on resistance thermometry) was also fabricated underneath the combustion electrodes to measure the temperature changes obtained. A signi cant amount of bubble production was seen up to 30 kHz but after that the bubble production decreased drastically, although the amount of faradaic current measured remained unchanged, signifying combustion. The temperature changes measured were also found to increase above this threshold frequency of 30 kHz.

Next, non-reactive molecular dynamic simulations were performed to determine as how does the surface tension of water surrounding the electrodes is affected by the presence of dissolved external gases, which would in turn help to predict the pressures inside nanobubbles. Knowing the bubble pressure is a perquisite towards understanding the combustion process. The surface tension of water was found to decrease with an increase in the supersaturation ratio (or an increase in the external gas concentration), thus, the internal pressure inside a nanobubble is much smaller than what would have been predicted using the planar-interface surface tension value of water. Once the pressure behavior as a function of external gas supersaturation was understood, then as a next step, reactive molecular dynamic simulations were performed to study the effects of surface-assisted dissociation of H2 and O2 gases and initial system pressure on the ignition and reaction kinetics of the H2/O2 system at nano-scales. A signi cant amount of hydrogen peroxide (H2O2), 6-140 times water (H2O), was observed in the combustion products. This was attributed to the low temperature(~300 K) and high pressure (2-80 atm) conditions at which the chemical reactions were taking place. Moreover, the rate at which heat was being lost from the combustion chamber (nanobubble) was also compared to the rate at which heat was being released from the chemical reactions and only a slight rise in the reaction temperature was observed (~68 K), signifying that, at such small-scales, heat losses dominate.