STATUS AND PERSPECTIVES OF LASER IGNITION OF A CRYOGENIC RESEARCH RCS THRUSTER

Transcription

1 STATUS AND PERSPECTIVES OF LASER IGNITION OF A CRYOGENIC RESEARCH RCS THRUSTER Michael Börner and Chiara Manfletti Institute of Space Propulsion, German Aerospace Center (DLR) June 3, Abstract 2 Introduction to laser ignition The ignition of cryogenic rocket engines is a critical step towards a reliable functioning of the engine. This process is characterized by the injector head geometry, the ignition method, the thermodynamical condition of the injected propellants and the sequencing of the propellant injection. This article reports on the advances towards the laser ignition of a cryogenic research reaction control system (RCS) thruster with 5 coaxial injectors. In the last years, laser ignition of a single injector configuration has been investigated with hundreds of ignitions performed for in vacuum injection and the two propellant combinations LOx/CH 4 and LOx/H 2 at the test bench M3.1 of the DLR Institute for Space Propulsion [11, 12]. For future applications of laser ignition in upper stage launcher engines or RCS engines, a broader background knowledge of the laser ignition processes of multi-injector configurations is needed but which is not available today. An injector head consisting of five individual but identical coaxial injectors has been designed and qualified for this purpose. In this configuration, one injector is surrounded by four injectors for boundary effect shielding. The combustion chamber used for the single injector configuration tests is foreseen for the upcoming tests and is optically accessible from three sides. Multiple optical diagnostics like high-speed Schlieren, highspeed OH* imaging and laser induced breakdown spectroscopy have been installed at the test bench and can be operated simultaneously. Cold flow checks with LOx and gaseous H 2 have recently been carried out and are presented in this paper. Finally, a brief outlook is given on further laser ignition research activities at the test bench M3.1. Key words: laser ignition, cryogenic propellants, green propellants, RCS engine, rocket engines michael.boerner@dlr.de More than ten years ago the search for replacements of hypergolic but toxic propellants was initiated. In particular for RCS engines cryogenic propellants like LOx/LH 2 or LOx/GH 2 are of great interest due to their availability on launcher systems and to their high specific impulse [5]. For these propellant mixtures a reliable ignition system with little additional mass and simplicity is needed. As a result of the search for possible ignition systems, laser initiated ignition ( laser ignition ) is considered as one potential candidate. It has a set of advantages over other ignition techniques like spark plugs or catalytic methods: laser ignition provides precise timing of the ignition in the order of milliseconds, an extreme precise localization of the ignition energy and a higher ignition probability for different mixture ratios [7]. Furthermore, laser ignition allows nearly infinite re-ignitions of the engine. It does not require premixing of the propellants as ignition can be performed directly in the combustion chamber. Therefore no additional ignition system propellant valves are needed and no additional tubes and fluid control systems have to be installed. The technological developments in diode pumped solid state lasers have resulted in miniaturized and reliable laser systems, that provide enough laser pulse energy for possible applications in liquid rocket engines and are by now commercially available [8, 13]. In addition, laser ignition is not only limited to small scale engines. For the reliable re-ignition of non-hypergolic liquid upper stage engines of launchers, laser ignition is a possible candidate [9]. The steps towards laser ignition are: Focusing of a laser pulse carrying enough energy for reaching photon fluxes of I f ocus > W/cm 2, optical breakdown of the media to be ignited (a so-called plasma kernel ), heating of the plasma kernel by the rest of the laser pulse, transformation of the plasma kernel into the flame kernel (a localized combustion zone) and finally the development 1

2 of the flame kernel into macroscopic combustion thus ignition of the engine. These critical steps depend on the local thermodynamical conditions of the media and its flow conditions. In particular laser energy losses and flame kernel heat dissipation have to be considered to guarantee ignition [15]. Two fundamental technical laser ignition concepts have been performed so far: laser plasma driven ignition (LPI) and laser ablation driven ignition (LAI). These differ in essential characteristics: LPI implies a direct optical breakdown of the propellant 4. In the case of LAI, the laser pulse is focused onto a metal surface, that can be a dedicated metal target, the combustion chamber wall or the injector faceplate. The interaction of the target material with the laser pulse leads to lower laser pulse energies needed for creating a plasma in comparison to LPI. While LPI is applicable to all optically accessible regions within the combustion chamber, LAI is limited to those areas where the target can be mechanically placed. The most important disadvantage of LAI is the possible heat load on the target material caused by the combustion process. On the other hand, LPI implies higher minimum laser pulse energies that are necessary for reliable ignition (MPE) than LAI. The MPE level has been investigated during a high number of hot run tests at the test bench M3.1 in function of the propellants and ignition location [11]: For GH 2 /LOx this value has been found to be 14.5 mj for LAI and 72 mj for LPI. During the corresponding tests, the LOx feed pressure was 2.5 bar and the LOx temperature was 90 K while the H 2 was injected at 12 bar at 224 K. For different propellant injection conditions and pressures this MPE can change dramatically as shown in section 4. Many further parameter influence the MPE such as laser wavelength, the focusing elements, the laser pulse length, the thermodynamical properties of the propellant [16] and the location of the ignition. Especially the location of the ignition has a significant influence on the ignition probability for liquid rocket engines. The findings of recent research activities for LPI in a single coaxial injector RCS rocket engine are summarized in Fig. 1 taken from [11]. It was shown, that 100% ignition probability can be reached within the shear layer of the coaxial injected propellant and LOx as well as further down stream from the injector face plate. In future, laser ignition of multi-injector configurations shall be investigated at M3.1. In order to replace the single injector configuration used so far, a new modular 5 coaxial injector head has been designed and qualified. Its characteristics are described in section 3. In the same section, the test bench and its diagnostics adapted Figure 1: Results from laser ignition tests of a research 200N RCS LOx/GH 2 and LOx/GCH4 engine realized at the test bench M3.1 [10] to laser ignition tests are briefly described. The subsequent section 4 is dedicated to laser side aspects of the ignition process. Its understanding and the characterization of the laser system used for ignition tests is fundamental for a profound evaluation of the ignition tests. In section 5, data of the first cold flow checks of the multi injector configuration are presented. A short summary and an outlook on the upcoming tests are given in the last section. 3 Experimental set-up for M3.1 laser ignition tests The M3.1 test bench allows the testing of sub-scale liquid rocket engines and provides LOx at LN 2 - temperature as well as propellants like H 2 at ambient or LN 2 -temperature and CH 4 at ambient temperature. LOx is cooled by means of a heat exchanger driven by liquid nitrogen. The injector head is in thermal contact with LN2 for short thermal transients after propellant valve-opening thus minimizing two-phase injection of the LOx into the combustion chamber. The test bench has a vacuum facility for setting different levels of chamber pressures before a test run. It consists of a vacuum tank (1500 liters) and an ejector system consisting of a single-stage, oil-sealed rotary vane pump and an interface to the M3.1 test chambers. The realizable pressures range from 80 mbar to ambient. For future RCS tests low pressure levels before propellant injection are necessary to represent in-space injection conditions and allow the investigation of thermodynamical effects like flashing of the LOx and Mach-level injection of the propellant like GH 2 which occur during the transient injection phase [12]. The general optical set-up of the test bench is shown in figure 2. It consists of a high-speed Schlieren set-up and 2

3 OH*-camera top view on combustion chamber light source tioned above, two sides are as well equipped with polished quartz windows for optical diagnostics. The injec- mirror #2 Schlieren edge high speed Schlieren camera flow direction mirror #1 spectrometer Figure 2: Set-up of the optical diagnostics a simultaneous high-speed OH*-imaging set-up. Furthermore, a spectroscopy alignment has been installed for laser induced breakdown spectroscopy of the laser plasma kernel. The high-speed Schlieren set-up is realized in a standard Z-configuration for minimal optical aberration and due to space saving aspects. The high speed camera (Fastcam SA-X) was set to frames per second (fps) for all tests presented in this report with a resolution of 1024 x 368 pixels. Optical access to about 75% of the combustion chamber is provided by its two lateral windows. As the Schlieren set-up is perpendicular to the direction of propellant injection, a detailed analysis of the transient injection phase, the ignition process and the flame development can be carried out. In order to ensure the same field of view for the OH*- diagnostics, a 20 x 10 cm 2 coated glass plate (mirror #2) was introduced into the Schlieren set-up: The coating is high-reflecting for nm and reflects the light of the OH* radicals (A 2 Σ, ν = 0 = X 2 Π, ν = 0: nm [4, 14]) out of the Schlieren system towards the OH* camera. To avoid any background distortion from the light source an identical mirror (mirror #1) was positioned on the other side of the combustion chamber to filter light of nm. 3.1 Description of the injector head, combustion chamber and laser system The combustion chamber used for the cold flow checks presented in this paper is the same as used for the single injector laser tests. This guarantees a maximum of comparability between the single injector configuration and the multi injector configuration. The chamber is a cylindrical combustion chamber with diameter compensation of window-induced contraction. The upper side has a small and exchangeable polished quartz window for optical access for the laser beam (see Fig. 3). As men- Figure 3: Front view on the injector face plate mounted into the combustion chamber. From the top, the laser beam can be coupled in. Optical diagnostics are realised via the left and right windows. tor head is a 5 coaxial injector head with LOx core and coaxial propellant injectors. One central injector is surrounded by 4 injectors equally spaced around the central one to shield it from boundary effects like recirculation of the propellants. The angular orientation of the four outer injectors is realized for optimal Schlieren and OH* diagnostics access to the central injector. Furthermore the free optical access to the central injector and to the two upper radial injectors for the ignition laser is guaranteed (see Fig. 3). The injector head is designed to deliver the same total mass flow as the previously used single injector head and all 5 injectors are meant to have 20 % of the original single injector mass flow. In this way, a comparison between the two injector pattern can be realized by keeping the mass flows constant, but increasing the number of injectors. The laser system (Quantel YG981 E10) that will be used for the ignition tests is a flashlamp-pumped q-switched Nd:YAG laser. The laser pulse is frequency doubled to 532 nm for better focusing via a plano-convex coated lens (f =75mm) and has a pulse length of about 9 ns. The beam diameter is about 8 mm. The laser pulse energy is measured via a beam splitter and a power-meter and the laser pulse form is verified via a fast photodiode. In Fig. 4 the focused laser pulse within the combustion chamber at atmospheric pressure is shown. The fact that laser light is scattered from the laser plasma, visible by 3

4 Figure 4: Laser plasma within the combustion chamber in front of the injector face plate Figure 5: Intensity thresholds for plasma generation in function of combustion chamber pressure. The laser parameter are:λ = 532 nm, f = 75 mm, d beam = 8 mm. The uncertainty of the determined intensities is about 10 9 W/cm 2 illumination of the injector faceplate, shows that a part of the laser pulse energy is not heating the plasma and is therefor lost for the ignition process. Up to 90% of the laser pulse is not converted into the laser plasma in unfavorable conditions [15]. This underlines the necessity for a profound understanding of the laser plasma interaction, single laser pulse diagnostics like pulse form and energy and the characterization of the laser system used for ignition tests. Some aspects of this topic are presented in the following section. 4 Laser-plasma aspects of laser ignition The minimum laser pulse energy needed for ignition (MPE) is one of the most critical parameters for the laser ignition of rocket engines. The lower this value is, the smaller the laser system has to be to deliver the needed pulse energy. This directly influences the over-all weight of the laser ignition system. Although the miniaturization of such systems has made progress, the MPE is still the dominating parameter to be minimized. For optical breakdown a certain threshold intensity at the focal point has to be reached. The focal intensity can be calculated by I f ocus = E pulse /τ/a f ocus where I f ocus is the intensity at the focal point of the focusing lens, E pulse the laser pulse energy, τ the FWHM laser pulse length and A f ocus the focal spot area calculated by the Gaussian beam approximation. As τ and A f ocus are fixed, the focal point intensity can only be modified by the laser pulse energy. In general, the threshold intensity for plasma formation can be written as a power law of the pressure of the medium in which ignition has to be realized: I thr p k with k [0; 1]. This dependency has been determined by various authors and depends in particular on the laser wavelength and τ [15]. To characterize the optical system of the focusing lens and the upper combus- tion chamber window, I thr has been measured in function of the combustion chamber pressure in N 2. The results are shown in Fig. 5: the lower, blue line represents the first occurrence of plasma as the laser pulse energy is increased for a given pressure within the combustion chamber. The black, upper line represents the threshold for 100% plasma formation probability. As the optical breakdown is a statistical process, a spectrometer was used to record the breakdown spectra of the plasma for 100 laser shots to determine weather breakdown occurred or not for a given laser pulse energy and pressure level. It shows, that the threshold intensity increases with lower pressure levels according to the power law I thr p k with k = This general observation is of importance for in-space application of laser ignition as the intensity threshold values, thus the MPE, for a certain test configuration cannot simply scaled to lower pressure levels. As mentioned above, not all of the laser pulse energy is converted into the laser-plasma. The leading edge of the laser pulse, when considered as a Gaussian profile, is transmitted without creating a plasma as the threshold intensity is not yet reached: I(t < t thr ) < I(t thr ). Once the threshold intensity is reached, plasma is formed and nearly all laser energy is absorbed by the plasma due to inverse bremsstrahlung. The lower the laser beam energy is and the smoother the leading edge of the temporal beam profile is, the more energy is lost for the ignition process [3]. In Fig. 6, the absorbed laser-plasma energy normalized to the laser beam energy is shown in function of the laser pulse energy normalized to the threshold laser pulse energy. It shows that for a multiple of the threshold laser energy for plasma creation, the absorbed pulse energy increases non-linear. Thus for laser ignition 4

5 Figure 6: Plasma energy normalized to laser pulse energy in function of laser beam energy normalized to the plasma creation threshold in air at atmospheric conditions. The uncertainty of the measured laser pulse energy is about 0.5 mj. Figure 7: Laser-plasma and the corresponding sonic wave emerging radially from the focal spot at a speed of 358 ± 15m/s. a steep temporal laser profile is favorable to avoid transmission of the leading edge of the laser pulse. Furthermore, MPE values are only valid for a given laser pulse length as they vary for modified temporal laser pulse characteristics. Further energy dissipation mechanisms during laser ignition are radiation losses of the plasma and shock wave generation (see Fig.7). According to [2], the shock wave transforms into a sonic wave, which has been verified by high speed Schlieren measurements. In Fig.7 the sonic wave is clearly visible. The velocity of the sonic wave was determined by the simple distance-time relation as 358 ± 15m/s which is the speed of sound in air (c 0 = 343 m/s). In order to learn more about the spectral characteristics of the location of laser ignition, a spectrometer set-up has been installed at the test bench. Via one of the lateral windows, optical access to the optical breakdown area is possible. It is envisaged to perform laser induced breakdown spectroscopy (LIBS) during laser ignition [17, 18]. By an optical alignment, the LIBS area is focused into the spectrometer. To test the set-up, LIBS in air under atmospheric conditions has been realized for laser pulses energies of 20 mj ±1mJ. Spectras with several delays to the maximum of the laser pulse intensity have been taken. They are shown in Fig. 8. For the first two spectra, the laser pulse at 532 nm is clearly visible, followed by a broad excitation due to the hot laser plasma, a region of high temperatures of about 10 5 to 10 6 K. The higher the delay to the laser pulse is, the clearer the transition lines of singly ionised nitrogen occur (λ NII = nm, λ NII = nm, λ NII = nm). Figure 8: Consecutive spectra of laser induced breakdown in air at standard atmosphere. 5 Cold flow tests In order to prepare the laser ignition tests, cold flow tests with LOx and H 2 have been performed. Especially the transient propellant injection behavior of the 5 injector configuration has been investigated into. For this purpose LOx and H 2 have been injected independently from one another. The Schlieren high speed set-up as well as static and dynamic pressure sensors mounted in the injector head and combustion chamber have been used to analyze the injection characteristics. The LOx was injected at 105 K, while H 2 was injected at 285 K. The corresponding injection conditions of the propellants are given in table 5. In Fig.9 the LOx injection is shown. The upper two pic- 5

6 H 2 LOx density [kg/m 3 ] 0.08 ± ± 50 velocity [m/s] 3440 ± ± 0.5 mass flow [g/s] 8 ± ± 5 temperature [K] 282 ± ± 5 Table 1: Injection conditions for the cold flow tests Figure 10: Time-averaged picture for the complete H 2 injection phase. proves the coaxial injection of the LOx for all injectors as well as the identical averaged spray characteristics. The same procedure has been applied to the H 2 injection test and the arithmetically time-averaged image is shown in Fig. 10. Although the hydrogen has been injected into sea level pressure conditions, Mach-level injection can be observed. This is due to the pressure drop from the H 2 dome pressure of about 10 bar to the pre-injection pressure of 1 bar in the chamber. As these cold flow tests are just the first of a series, further tests have to be performed to find the best sequencing for the propellant injection for laser ignition tests. 6 Summary and outlook Figure 9: 3 high speed pictures of LOx injection at times of t s (upper image) and t s (middle image). The lower image is a time-averaged image for the whole transient LOx injection phase. tures show the LOx injection at times of t s (upper image) and t s (middle image), where t 0 is the projected time of laser ignition in future ignition tests. In the upper image, the pre-injection of the LOx through the central injector can be seen. The delay of the outer injectors with reference to the central injector is 6±2 milliseconds. As the central injector is meant to be ignited via LPI or LAI, this pre-flow is a design feature of the injector head. The middle image shows the LOx injection through all injectors at t s: The LOx core is clearly visible and the coaxial injection of all 5 injectors can be verified. As the Schlieren set-up is a projection to a 2D plan, the 2 upper injectors and the 2 lower injectors overlay onto the images (see Fig. 3). Once again, the coaxial injection can be verified by the fact that the upper row of 2 injectors show the same shadow pattern as the single central injector. The lower image is a simple arithmetically timeaveraged image of all images of the LOx injection and In this conference contribution, aspects of the laser ignition research at the DLR Institute of Space Propulsion have been presented. In particular, steps towards a multi-injection configuration and laser-plasma aspects have been reported. In future, further cold flow checks will be performed to characterize the injection characteristics for sea level and in-space like configurations. Finally, LAI and LPI tests will be carried out for the described 5 coaxial injector configuration. The Schlieren and OH* high speed diagnostics will be used to investigate into the flame development. A comparison with the single injector configuration will show the influence of the additional outer injectors on the ignition process and probability. A better understanding of the ignition transients and flame development for multi-injector configurations for in-space conditions will serve as a basis for further development in the area of laser ignition of rocket engines. Furthermore, the miniaturization of diodepumped solid-state lasers continues and pulsed fiber laser systems increase the laser pulse output energy to levels close to the MPE for laser ignition applications [1, 6]. This development opens a wide field of applications in the space propulsion sector in which weight is a sensitive parameter. 6

7 7 Acknowledgments The authors acknowledge the support of Johann Fröse, Markus Dengler and Michael Zepmeisel for their assistance at the M3.1 test bench. References [1] B. Beaudou, F. Gerôme, Y. Y. Wang, M. Alharbi, T. D. Bradley, G. Humbert, J.-L. Auguste, J.-M. Blondy, and F. Benabid. Millijoule laser pulse delivery for spark ignition through kagome hollowcore fiber. Opt. Lett., 37(9): , May [2] Harold L. Brode. Numerical solutions of spherical blast waves. J. Appl. Phys., 26(6): , [3] Y.-L Chen, J.W.L Lewis, and C Parigger. Spatial and temporal profiles of pulsed laser-induced air plasma emissions. Journal of Quantitative Spectroscopy and Radiative Transfer, 67(2):91 103, [4] G.H. Dieke and H.M. Crosswhite. The ultraviolet bands of OH Fundamental data. Journal of Quantitative Spectroscopy and Radiative Transfer, 2(2):97 199, [5] Keichi Hasegawa and Masahiro Sato. Laser Ignition Characteristics of GOX/GH2 and GOX/GCH4 Propellants. AIAA, , [6] Sachin Joshi, Nick Wilvert, and Azer P. Yalin. Delivery of high intensity beams with large clad stepindex fibers for engine ignition. Appl. Phys. B, 108, [7] H. Kopecek, M. Lackner, J. Klausner, M. Weinrotter, E. Wintner, G. Herdin, C. Forsich, S. Charareh, and F. Winter. Laser ignition of methane-air mixtures at high pressures and diagnostics. Journal of Engineering for Gas Turbines and Power, 127: , [11] Chiara Manfletti. Laser ignition of an experimental cryogenic reaction and control thruster: Ignition energies. Journal of Propulsion and Power, [12] Chiara Manfletti. Laser ignition of an experimental cryogenic reaction and control thruster: Preignition conditions. Journal of Propulsion and Power, [13] Chiara Manfletti and Gerhard Kroupa. Laser ignition of a cryogenic thruster using a miniaturised nd:yag laser. Opt. Express, 21(S6):A1126 A1139, Nov [14] S Pellerin, J M Cormier, F Richard, K Musiol, and J Chapelle. A spectroscopic diagnostic method using uv oh band spectrum. Journal of Physics D: Applied Physics, 29(3):726, [15] Tran X. Phuoc. Laser-induced spark ignition fundamental and applications. Optics and Lasers in Engineering, 44(5): , [16] Johannes Tauer, Heinrich Kofler, and Ernst Wintner. Laser-initiated ignition. Laser & Photon. Rev., 4(1):99 122, [17] L. Zimmer, K. Okai, and Y. Kurosawa. Combined laser induced ignition and plasma spectroscopy: Fundamentals and application to a hydrogen air combustor. Spectrochimica Acta Part B: Atomic Spectroscopy, 62(12): , ce:title A Collection of Papers Presented at the 4th International Conference on Laser Induced Plasma Spectroscopy and Applications (LIBS 2006) /ce:title. [18] Laurent Zimmer and Shigeru Tachibana. Laser induced plasma spectroscopy for local equivalence ratio measurements in an oscillating combustion environment. Proceedings of the Combustion Institute, 31(1): , [8] Gerhard Kroupa, Georg Franz, and Ernst Winkelhofer. Novel miniaturized high-energy nd-yag laser for spark ignition in internal combustion engines. Optical Engineering, 48, [9] Larry C. Liou. Laser ignition in liquid rocket engines. AlAA , [10] Chiara Manfletti. Laser Ignition of a Research 200N RCS LOx/GH2 and LOx/GCH4 Engine. In 48th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit,