발간 현황

A small quantity of ADN was synthesized in a laboratory-scale experiment at Kyunghee University in Korea and SEM image is shown in Fig. 1[9]. The propellant contains 63% by weight of ADN as oxidizer, 14% of deionized water, 18.4% of CH3OH as fuel, and 4.6% of NH3 used as stabilizer. The above composition follows the research conducted by Wingborg et al., and was previously shown to comply with detonability, storage, toxicity and shock sensitivity requirements from the Swedish Defense Agency (FOI)[10]. Another propellant, the FLP-106, was studied in the framework of the FP7[5]. The properties of the propellants are shown in Table 1[1]. Both propellants have zero oxygen balance for maximum performance, low shock sensitivity and are compatible with the materials used in conventional propulsion systems whereas the HAN-based propellant, AF-M315E, requires titanium components[2]. The FLP-106 has a sensibly lower volatility and higher density compared to LMP-103S making it slightly more performant[11].

Fig. 1 
SEM image of crystallized ADN[9].

The propellant was evaluated using DSC-TG analysis which evaluates the enthalpy changes as well as its decomposition profile, as illustrated in Fig. 2. The thermal decomposition of the propellant shows a different behavior compared to aqueous ADN with a single broad peak at 175℃ due to the inhibiting nature of methanol as discussed previously[5]. The decrease of the sample’s mass is due to the evaporation of the gaseous species and corresponds to the exothermic decomposition reaction observed by DSC. The experiment was performed without a cap to minimize the risk of explosion.

Fig. 2 
Thermal decomposition by DSC-TG.

Commercial Alumina pellets were obtained from Alfa-Aesar, were crushed and then sieved to an 18-20 mesh size. Subsequently, the pellets were impregnated with a precursor solution of Lanthanum, dried and calcinated to form a hexaaluminate support. As a doping agent, Lanthanum will limit the crystallization of alumina at higher temperatures[12]. The impregnation of the active material is performed by conventional wet impregnation. A bimetallic catalyst was selected to decrease the ignition temperature of the propellant, following the analysis of the catalytic decomposition of 50 wt% aqueous ADN by Kamal[6]. The catalyst support, LHA, is first impregnated with Platinum, calcinated and then impregnated with Copper without intermediate reduction. Finally, the catalyst is further calcinated and reduced (Fig. 4). Through this procedure, a portion of the active material may consist in an alloy of platinum and copper. For increased safety, only 5 L of the propellant combined with 8 mg of crushed catalyst were analyzed by DSC-TG. The sample was heated from 25°C to 250°C at a rate of 5 °/min, following an isotherm of 5 min at the starting temperature. According to the results displayed in Fig. 3, the catalytic decomposition temperature occurs at a lower value (129℃) when using a Pt-Cu catalyst than for the thermal decomposition or catalytic decomposition of the propellant using a monometallic Pt catalyst (175℃).

Fig. 3 
Catalytic decomposition by DSC-TG.

Fig. 4 
18-20 Mesh Pt-Cu/LHA catalyst.

Furthermore, the sharp peak demonstrates excellent catalytic reactivity and higher decomposition reaction rate.

Subsequently, specific surface area (BET) and pore analyses reveal that the porous structure of LHA is sufficiently maintained after a calcination of 5h at 1200℃, with a specific surface area of 24.5 m²/g (Table 2)[4]. SEM analysis was used to evaluate the active material loading, distribution and crystallite size on the catalyst pellets. Five pellets were selected, and EDS point analysis over an area of 200 μm² gave an average loading for Pt of 4.3% which is much lower than the corresponding weight balance approximation giving 13.7% by weight. The discrepancy between these results may be explained in part by the error involved in EDS analysis which lies around 15% for the elements studied, or by the impossibility of detecting metallic particles lodged inside the pores. Popa shows that Platinum crystallites smaller than 20 nm can be obtained using the same impregnation procedure[13]. The Secondary Electron image below shows white crystals which are larger than the pores of the LHA support (Fig. 5). Platinum is easily distinguished from the other elements due to its higher atomic number, but EDS point analysis reveals that the white particles sometimes consist of an alloy of Cu and Pt. For analyzing the nano-sized Pt crystallites in the pores, however, Transmission Electron Miscroscopy analysis should be considered.

Fig. 5 
Surface image of the catalyst.

The thruster was designed to produce a nominal thrust of 5 N in space, however, sea-level testing imposed a nozzle expansion ratio of 2.16 resulting in a thrust of 3.6 N. With the help of NASA’s CEA and the enthalpy of formation Fig. 3 Catalytic decomposition by DSC-TG. of aqueous ADN and ammonia from the literature, the propellant properties were obtained in Table 1[14]. The design mass flow rate was 2.08 g/s for the propellant, assuming frozen composition in the nozzle. Sensys PSHD series pressure sensors were used to measure steady, static pressure at two points in the catalyst bed and in the chamber as illustrated in Fig. 6.

Fig. 6 
5 N scale thruster design.

Due to the small thruster scale and the minimal diameter of more heat-resistant R-type thermocouples, smaller K-type thermocouples were employed to measure the temperature in the catalyst bed. Nevertheless, since the K-type has a temperature limitation of around 1250 ℃, only the first thermocouple was left in place. The feeding system was simplified to accommodate the small quantity of synthesized propellant, as a result of which, no drain valve, safety check valve and filter were added after the tank, as shown in Fig. 7.

Fig. 7 
Feeding system and thruster.

Firing tests were conducted for 3 seconds using a feeding pressure of 12.5 bar considering estimated pressure drops in the catalyst bed, across the injector and the orifice-type mass flow meter. The short firing time is a result of the thruster material limitations and the quantity of available ADN-based propellant.

The first firing test resulted in a contained explosion in the catalyst bed characterized by a pressure spike in the thruster above the feeding pressure as portrayed in the pressure curve in Fig. 8.

Fig. 8 
Preliminary firing test – Pressure.

Since the inspection of the thruster and feeding system did not result in any noticeable damage to the equipment, two further firing tests were conducted, leading to the pressure and temperature graphs below (Figs. 9, 10). Measured mass flow rate was 1.13 g/s, average chamber pressure was 3 bar, and the C^* efficiency was 53.6%. The sudden increase in mass flow rate at the beginning of the test is related to the high initial pressure difference across the solenoid valve. A cavitating Venturi valve could be used to regulate the mass flow rate and reduce the risk of hard start, which was experienced here. The pressure curve shows that the mass flow rate was well below the design value, at about 1.14 g/s and so was the chamber pressure, at 3 bar. Nevertheless, the elevated temperature measured by the thermocouple located at the beginning of the catalyst bed indicates proper decomposition of the ADN-based propellant as shown in Fig. 10. Furthermore, the thruster operation was also characterized by an invisible plume: Fig. 11 shows the thruster in operation with visible pressure sensors and pre-heating apparatus but invisible exhaust plume due to the exhaust products mainly being carbon dioxide, nitrogen and nitrous compounds.

Fig. 9 
Firing test – Pressure.

Fig. 10 
Firing test – Temperature.

Fig. 11 
Thruster operation.

Examination of the catalyst bed and distributor after the experiment revealed major damages caused by the explosion during the first test, as portrayed in the picture of Fig. 12. The decomposed propellant flow caused the crack inside the damages shown in the Fig. 12, occurred chamber pressure increase during the firing test shown in Fig. 9. The explosion may have been caused by the sudden expansion of air bubbles trapped in the feeding line due to the absence of a bleeding valve. The pressure drop across the injector which stands at a very low 6% of the chamber pressure, may be another cause for instability leading to a hard start. The mass of crushed catalyst pellets on the distributor explains the significant pressure drop from the catalyst bed to the chamber. A better thruster design will have a longer combustion chamber to reduce the thermal and mechanical stress on the pellets and allow the full combustion of the propellant.

Fig. 12 
Post-firing thruster condition.

The catalyst was characterized once more following the firing tests to examine its thermal stability and the properties can be found in Table 2 below. The specific surface area after the firing test was half of the value prior to the test. By SEM analysis, no significant change in the crystallite size or dispersion was observed, although consistently lower Pt weight percent was identified on the crushed catalyst. These changes could also have resulted from the mechanical stresses generated by the explosion, owing to a maximal pressure of 30 bar in the catalyst bed. However, thermal effects are not excluded since crushing of the pellets in powder form was shown not to affect the nano-porous structure. The picture of the distributor plate in Fig. 12 shows crushed LHA catalyst stuck to the distributor plate.