Mass spectrometric study of combustion of GAP- and ADN-based propellants

Mass spectrometric study of combustion of GAP- and ADN-based propellants

Mass Spectrometric Study of Combustion of GAP- and ADN-based Propellants L. V. KUIBIDA, O. P. KOROBEINICHEV,* A. G. SHMAKOV, E. N. VOLKOV, and A. A. P...

298KB Sizes 2 Downloads 53 Views

Mass Spectrometric Study of Combustion of GAP- and ADN-based Propellants L. V. KUIBIDA, O. P. KOROBEINICHEV,* A. G. SHMAKOV, E. N. VOLKOV, and A. A. PALETSKY

Institute of Chemical Kinetics and Combustion, Russian Academy of Science, Novosibirsk 630090, Russia The flame structure of composite propellants and sandwiches based on ammonium dinitramide (ADN) and glycidyl azide polymer at 0.015 to 0.3 MPa was studied by molecular beam mass spectrometry. A zone near the surface, ⬃1.5 mm wide, was detected, where reactions occur. The gas composition near the surface of burning ADN laminae at 0.1 MPa was close to that near the surface of burning pure ADN at 0.3 MPa. Among the species responsible for reactions in the flame near the surface, the most probable are HNO3, dinitraminic acid, and the vapor of ADN. The luminous zone of the flame extends more than 10 mm from the surface. The composition of the final combustion products has been determined by freezing at the temperature of liquid nitrogen and indicates incomplete combustion. The temperature profiles measured with thin thermocouples confirm the measured widths of the near-surface and luminous zones. The final temperature at the pressure of 0.3 MPa is as high as 2600 K. © 2001 by The Combustion Institute

INTRODUCTION Measurements of flame structure may be useful for elucidating the nature and the pattern of the combustion of solid rocket propellants (SRP). By flame structure we mean the spatial distribution of the temperature and concentrations of species in the combustion wave, including gasification products, intermediate and final combustion products. In the present paper, the flame structure of composite propellants and sandwiches based on the oxidizer ammonium dinitramide (ADN) and an active combustible binder glycidyl azide polymer (GAP) was studied. ADN and GAP are new prospective components [1, 2] of SRP. GAP has the structural formula H[-O-CH (-CH2-N3)-CH2-]nOH, and its characteristic peculiarity is the presence of an -N3 group in each monomer. Degradation of this group in the first stage of thermal decomposition of the polymer results in the release of ⬃376 kJ/mol of heat, which is the basis for the ballistic properties of GAP in the propellant [3]. In sandwiches, unlike composite propellants, the position of the oxidizer and the combustible are arranged in alternating laminae. They are not used in real SRPs, but are more convenient for studying flame structure and for computer simulations. Some studies of the combustion of *Corresponding author. E-mail: [email protected] COMBUSTION AND FLAME 126:1655–1661 (2001) © 2001 by The Combustion Institute Published by Elsevier Science Inc.

such sandwiches have been performed [4 – 6] (E. W. Price, School of Aerospace Engineering, Georgia Institute of Technology, Atlanta, 1998). In these, either a few alternating laminae of oxidizer and combustible, or only two oxidizer laminae with a combustible lamina between them were used. Apart from ballistic characteristics, such as burning rate and its dependence on pressure, the shape of the burning surface was studied either on extinguished samples, or during burning, by video filming. The flame structure of some sandwiches based on ammonium perchlorate at low pressures has been studied [4] using microprobe mass-spectrometry. Also the flame structure with sandwiches, consisting of two thick (⬃1.5 mm) laminae of ADN and one thick (0.3– 0.4 mm) GAP lamina between them, has been investigated [5]. Ignition and burning were initiated by laser heating with a power of 6.27 MJ/m2/s and visualization techniques with plane laser-induced fluorescence was used. Using a wavelength for the fluorescence of nitric oxide, the shape of the diffusion flame over the GAP lamina during ignition and later during stationary burning was studied. The presence of a dark zone in stationary burning (in the first 10 ms after the laser was switched on) was demonstrated. This zone extended to a distance of not less than 3 mm from the burning surface. The maximum fluorescence was achieved at a distances of 10 mm and more. The temperature 0010-2180/01/$–see front matter PII 0010-2180(01)00274-7

1656

L. V. KUIBIDA ET AL.

calculated from the rotational spectrum of NO was ⬃1450 K in the dark zone and 3200 K at its maximum. MATERIALS AND METHODS Ammonium dinitramide NH4(NO2)2 was synthesized at the Zelinsky Institute of Organic Chemistry of the Russian Academy of Sciences. It is a white powder consisting of needle-like crystals, with a wide range of size from 1 ␮m to 1 mm. The crystal’s density is 1.82 ⫻ 103 kg/m3; the melting point is 365 K. In solid rocket propellants ADN is used in the form of granules with a size from sub-millimeter to tens of microns [7]. We used GAP with a chain length of n ⫽ 20, synthesized at the St. Petersburg Technological University. It is a viscous yellow fluid with a specific mass of ⬃1.3 ⫻ 103 kg/m3. When the polymer is used in a real SRP, to obtain the necessary mechanical properties, it is converted to a rubberlike condition by the standard curing procedure [3]. The number of N3 groups is not thereby changed, but part of the terminal OH groups are replaced by NCO bridges. Cured GAP has the elemental composition C3.3H5.6O1.12N2.63 [3]. We used two types of sandwich with various lamina thicknesses. Sandwich 1 consisted of five 0.8-mm-thick ADN laminae and six 0.2-mmthick laminae of cured GAP. This corresponds to a composition of 82 wt.% ADN ⫹ 18 wt.% GAP. This is close to the stoichiometric composition (82.5% and 17.5%, respectively). The sample size was 5 ⫻ 6 ⫻ 8 mm. For preparation, 0.8 ⫻ 6 ⫻ 8 mm ADN plates and 0.2 ⫻ 6 ⫻ 8 mm cured GAP plates were used. ADN plates were pressed at a pressure of 700 MPa. To maintain the sample’s shape during burning, two 0.2 ⫻ 6 ⫻ 8 mm glass plates over the end GAP laminae were used. To prevent the spread of the flame to the side surface of the sample, these were protected with a perfluorinated lubricant (polymonochlortrifluoroethylene). Sandwich 2 consisted of three 1.6-mm ADN laminae, two 0.4-mm GAP laminae between them, and two another 0.2-mm GAP laminae at the ends. Studies of flame structure were done at 0.1 MPa in argon on an automated mass spectrometric set-up with molecular beam sampling based on a time-of-flight mass spectrometer

Fig. 1. Profile of the burning surface of a sandwich and its position with respect to the probe.

MCX-5 (Electromechanical factory B-2613, Sumy, USSR) [8]. The quartz probe had an orifice 0.1 mm in diameter. Ignition was performed with a flat electric spiral. The initial distance from the sample to the orifice was ⬃3 mm. Immediately after ignition, the step motor of the scanning mechanism was switched on to move the burning sample toward the probe at a velocity exceeding the rate of burning by 5 mm/s. Simultaneously with this, mass spectra were recorded by the time-of-flight mass spectrometer, and video filming was performed. For video filming, a Panasonic M3000 camera (Secaucus, NJ, USA) with a supplementary lens with f ⫽ 50 mm was used. The frame exposure time was 1/2000 s at a frame frequency of 25/s. Before the experiment, the sample was installed so that the probe touched the surface in the middle of the ADN lamina, as shown in Fig. 1. It is important to note that Fig. 1 was designed on the base of video filming of the burning surface of the sandwich. When the probe was touching a molten ADN lamina on the surface, a sharp jump in the pressure in the vacuum system and a sharp increase of practically all the peaks in a mass spectrum peaks occurred. Such a sharp increase of peak intensities is an indicator of the probe touching the surface of the burning sample. Knowing the time of this touching and the relative speed of convergence, one could calculate the distance from the burning surface for all the previously recorded mass spectra. Especially for thermocouple studies of the burning wave, samples of sandwich 1 were made in which the middle ADN lamina consisted of two plates of equal thickness (0.4 mm), between which thin thermocouples were placed. Two

GAP AND ADN-BASED PROPELLANTS

1657 TABLE 1

Characteristics of Combustion of Sandwiches at Various Pressures

Sandwich

P, MPa

Burning rate, mm/s

Maximal flame temperature, Tmax, K

Lamina thickness ADN/GAP, mm

Height of rising of GAP lamina over surface, mm

1 1 1 2

0.04 0.1 0.3 0.1

⬃1.3 3.2 ⫾ 0.2 ⬃8 3.3 ⫾ 0.1

no data 2100 ⫾ 100 2600 ⫾ 100 2150 ⫾ 40

0.8/0.2 0.8/0.2 0.8/0.2 1.6/0.4

no data 0.2–0.4 no data 0.4–0.6

thermocouples were made. One was of platinum-rhodium, welded from wires 50 ␮m in diameter and rolled to a cross-section of 15 ⫻ 140 ␮m. The other was of tungsten-rhenium, welded from wires with a diameter of 30 ␮m and rolled to a section of 15 ⫻ 60 ␮m. The thermocouples had no protective coating, and were burnt when they went out of sample into the flame. RESULTS AND DISCUSSION According to video pictures, on the burning surface of sandwiches, the GAP lamina rise over the ADN lamina by 0.4 to 0.6 mm. Table 1 presents the combustion characteristics for sandwich samples in an argon atmosphere: burning rate, maximal flame temperature and height of rising of GAP lamina over ADN at the burning surface. The burning rate of pure ADN compacted to a density of 1.79 ⫻ 103 kg/m3 is [9] 3.44 ⫾ 0.05 and 11 ⫾ 1 mm/s at pressures of 0.1 and 0.3 MPa, respectively (in argon). The maximal flame temperature increased with pressure rise. Figure 2 presents the change of intensity of some mass spectral peaks, depending on the distance from the burning sample for sandwich 1. One can see that near the burning surface there is a zone of ⬃1.5 mm where considerable changes of the following mass peaks take place: 46 (NO2 and HNO3), 18 (H2O), 30 (NO, HNO2, NO2 and HNO3), 44 (CO2 and N2O), 28 (N2, CO), 16 and 17 (NH3), and of weak peaks of 27 and 29 (HCN and CH2O). The main contribution to the change of peaks at 46, 44, 30, 17, and 16 near the burning surface of ADN has to be made by vapors [9] of ADN (ADNv) and dinitraminic acid. When the thickness of the lamina was dou-

bled (sandwich 2), the width of the surface zone did not change. However, the ratio of massspectral intensities near the burning surface approached the ratio characteristic of burning pure ADN at 0.3 MPa [9]. The results are presented in Fig. 3. In Table 2, mass peak ratios near the burning surface of sandwich 2 (in the middle of an ADN lamina) at 0.1 MPa and of pure ADN at 0.3 MPa presented. In order to estimate these ratios, the curves in Fig. 3 with strong pulsations were smoothed out. Mass intensities were normalized with respect to the peak at mass 46. The similarity of mass spectra presented in Table 2 indicates a similarity in the composition of the products near the surface of burning ADN for a sandwich (0.1 MPa) and pure ADN (0.3 MPa). As the treatment of mass-spectral data (in Table 2) for pure ADN at 0.3 MPa showed, there are the following species

Fig. 2. Profiles of mass spectral peak intensities when probing the thin-lamina sandwich 1 (0.8-mm ADN and 0.2-mm GAP) at p ⫽ 0.1 MPa.

1658

L. V. KUIBIDA ET AL.

Fig. 4. Temperature profiles obtained with thin thermocouples placed in the middle of the oxidizer lamina, and with a coated thermocouple installed at an initial distance of 1.5 mm for sandwich 1 at 0.1 MPa.

Fig. 3. Profiles of mass spectral peak intensities when probing the thick-lamina sandwich 2 (1.6-mm ADN and 0.4-mm GAP) at p ⫽ 0.1 MPa.

in the products near the burning surface (their mole fractions are indicated in brackets): ADN vapor (0.03), NH3 (0.08), NO (0.19), N2O (0.24), N2 (0.08), HNO3 (0.08), H2O (0.3). Analysis of Table 2 shows that, as for the combustion of pure ADN at 0.3 MPa, at the burning surface of ADN in a sandwich at 0.1 MPa, the main product responsible for the subsequent reactions in the flame zone is gaseous ADN. However, one can also see that, besides evaporation of ADN, reactions forming the above-mentioned products in the condensed phase take place. The presence of strongly TABLE 2 Ratios of Mass Peak Intensities near the Burning Surface of ADN in Sandwich at 0.1 MPa and of Pure ADN at 0.3 MPa Masses

46

30 44 28 17 16 18

Intensities, % 100 90 80 50 50 35 20 Pure ADN, p ⫽ 0.3 MPa [7] 100 86 44 20 47 ? 12

pulsating peak intensities for masses 16, 17, 30, 44, and 46, and the steadiness of masses 18 and 26 may be associated with the burning of GAP laminae. Video filming showed that during the combustion of GAP lamina, the formation of a carbonaceous carcass at the top of GAP lamina took place. At some time this carcass broke away. This process repeated itself and was possibly a reason for the pulsations. The same pulsations are characteristic of the luminous flame of the high-temperature zone. They correlate with pulsations in the peak intensities of mass spectra. In the case of sandwich 1 with thin laminae (0.8-mm ADN and 0.2-mm GAP), the pulsations are much less pronounced, and sometimes even invisible in video filming. Pulsations of temperature are also observed when measuring temperature profiles in the flame with a thin thermocouple. The results are presented in Fig. 4. The thermocouples demonstrated practically equal temperatures with strong pulsations in the gaseous phase. Also presented in Fig. 4 is a temperature profile measured with the W-Re thermocouple with a protective anticatalytic coating “Ceramobond569” (Aremco Products, Inc., Ossining, NY) [11], installed at an initial distance of 1.5 mm above the sample’s surface. One can see that these profiles coincide well at a distance of 1.5 to 2.5 mm. This confirms that the measured

GAP AND ADN-BASED PROPELLANTS temperature profiles are correct. The video filming showed a zone with a bright luminous flame at a distance of more than 5 mm from the burning sandwich’s surface. The thermocouple showed that the temperature there reached 2100 K at 0.1 MPa. Comparison of measurements for a sandwich flame (0.1 MPa) presented in Fig. 2 with analogous data (for a pure ADN flame at 0.3 MPa) on mass peak intensities also demonstrated their similarity. It consists of: 1) a drop in a 1.5 mm zone of the peak of mass 46 mainly responsible for ADN vapor; 2) incomplete reaction of NH3 (peaks of masses 16 and 17) at distance of 3 mm from the burning surface. At this distance, the measured temperature was 1000 to 1100 K (see Fig. 4). In the combustion of pure ADN at 0.6 MPa [9], a so-called second zone was observed where ammonia oxidation by nitric acid took place. Therein, a noticeable reaction of ammonia began at T ⬎ 1100 K. Comparison of these facts with the results obtained for a sandwich showed that with a sandwich the “second zone” was not apparent at distances of ⬍3 mm from the surface of burning ADN. In a similar way, the cured composite propellant containing 82.5 wt.% ADN and 17.5 wt.% GAP was studied. The sample’s diameter was 10 mm, its height was 10 mm, and the burning rate 2.75 ⫾ 0.05 mm/s at 0.1 MPa in argon. Visual observations of burning showed the absence of jets on the burning surface, as well as an absence of carbon core formation, and the presence of a pure transparent flame. By contrast, for ADN/HTPB propellants, flame jets appear and disappear on the burning surface, and after burning, a carbon core remains. The results of probing the flame of ADN/GAP-based composite propellant at 0.1 MPa are presented in Fig. 5. Here, just like the case of sandwiches, a zone was observed near the surface ⬃1.5 mm wide. By freezing and defrosting the combustion products, as used earlier for studying the composition of the products from burning GAP in an inert atmosphere [11], the final combustion products of sandwich 1 and of the mixture were analyzed; the results are in Table 3. The composition of products is not complete, because the amount of water was not controlled. The main products are N2 and CO2. In the final products of combustion of sandwich 1 and com-

1659

Fig. 5. Profiles of mass spectral peak intensities in probing a composite system at p ⫽ 0.1 MPa.

posite propellant (Table 3), NH3 and N2O are absent. There, the final burning temperature of sandwich 1 at 0.1 MPa is 2100 K (Table 1), which is close to the adiabatic temperature of pure ADN. In this way, in the combustion of sandwiches and of composite propellant at distances of ⬎3 mm from the burning surface, reactions in the second and third (N2O decomposition zone [9]) zones of pure ADN flame were completed. The flame structure with a composite propellant was also studied at low pressures. The combustion rate was 1.72 ⫾ 0.08 mm/s at a pressure of 0.04 MPa in argon and ⬃0.48 mm/s at 0.015 MPa. Mass 46, due to NO2, changed most with distance near the surface. The peaks TABLE 3 Composition of Final Combustion Products (Except Water) for ADN/GAP-based Systems in mg per g of Sample

Composite propellant Sandwich 1

N2

CO

CO2

H2

NO

292

34

246

3

15

214

⬍4

264

2

24

1660

L. V. KUIBIDA ET AL. CONCLUSIONS

Fig. 6. Behavior of mass 46 peak intensities near the burning surface of a composite system at various pressures.

of other masses (30, 44, 18, etc.) changed near the surface only slightly. Because at low pressures the sensitivity decreased considerably, their changes often turned out to be within the limits of noise. The behavior of the peak at mass 46 near the burning surface for various pressures is presented in Fig. 6. The intensity of this peak was normalized with respect to the intensity of the peak of mass 30 (NO), which was approximately constant throughout the flame zone. Figure 6 shows that mass 46 falls to zero at 1.5 to 2.0 mm from the surface, independent of pressure. At the lowest pressure presented (0.015 MPa), the peak of mass 46 did not fall to zero at distances of 1.5 to 4 mm. Therefore, in this case the reactions in the flame did not go to completion, so one could expect that in the combustion products there would be some intermediate compounds. In particular, the peak at mass 46 may be associated with HNO3, dinitraminic acid, and vapors of ADN. Unfortunately, it was not possible to derive profiles of concentrations of species from the intensities of peaks in a mass-spectrum. The main reason was the low accuracy of measuring weak peaks in a mass-spectrum; so it was not possible to identify species from their fragmentary peaks with certainty. Improved accuracy was prevented by the short time for which sampling was possible and also the unsteadiness of combustion.

Measurements on the flame structure and combustion characteristics of composite propellants and sandwiches based on the new prospective components ADN and GAP at pressures of 0.015 to 0.3 MPa have been obtained. The burning rates, temperature profiles, widths of combustion zones, and the heights to which GAP laminae rise over the burning surface of sandwiches are presented. Two zones in a flame were found: the first near the burning surface is a dark, low-temperature zone of width of 1.5 mm. This zone is similar to the first zone in a flame of pure ADN at pressure of 0.3 MPa, in which dissociation of ADN vapors to ammonia and dinitraminic acid, with subsequent decomposition of the latter, takes place. The second is a luminous zone with a width of 8 to 10 mm, where further oxidation of decomposition products of the sandwich’s components takes place. The information obtained on the chemical structure of combustion zones with sandwiches is qualitative. However, these data together with measurements on the combustion characteristics of sandwich and quantitative data on the composition of the combustion products of a composite propellant can be used for developing of a combustion model for composite propellants based on ADN and GAP. The work was supported by the U.S. government through Contract F61708-WO195. REFERENCES 1. 2.

3. 4.

5.

6. 7.

Frankel, M. B., Grant, L. R., and Flanagan, J. E., J. Propuls. Power 8:560 (1992). Schoyer, H. F. R., Schnorhk, A. J., Korting, P. A. O. G., van Lit, P. J., Mul, J. M., Gadiot, G. M. H. J. L., and Meulenbrugge, J. J., J. Propuls. Power 11:856 (1995). Kubota, N., and Sonobe, T., Propellants, Explosives, Pyrotechnics 13:172 (1988). Korobeinichev, O. P., Tereshchenko, A. G., Shvartsberg, V. M., Chernov, A. A., Zabolotnyi, A. E., and Emelyanov, I. D., Combust. Expl. Shock Waves 26:173 (1990). Parr, T., and Hanson—Parr, D., in Non-Intrusive Combustion Diagnostics (K. K. Kuo, and T. P. Parr, Eds.) Beggel House, New York, 1994, p. 571. Price, E. W., J. Propuls. Power 11:717 (1995). Ramaswamy, A. L., Combust. Expl. Shock Waves 36: 119 (2000).

GAP AND ADN-BASED PROPELLANTS 8.

9.

10.

Amosov, K. A., Kuibida, L. V., Korobeinichev, O. P, and Bolshova, T. A., in Flame Structure, Vol. 1 (O. P. Korobeinichev, Ed.), Nauka, Novosibirsk, 1991, p. 93. Korobeinichev, O. P., Kuibida, L. V., Paletsky, A. A., and Shmakov, A. G., J. Propuls. Power 14:991 (1998). Korobeinichev, O. P., Kuibida, L. V., Shmakov, A. G., and Paletsky, A. A., 37th AIAA Aerospace

1661

11.

Sciences Meeting and Exhibit, AIAA-99-0596, Reno, NV, 1999. Burton, K. A., Ladouceur, H. D, and Fleming, J. W., Combust. Sci. Technol. 81:141 (1992).

Received 17 October 2000; revised 11 May 2001; accepted 15 May 2001