Gravity effects on a gliding arc in four noble gases: from normal to hypergravity

Transcription

1 Plasma Sources Science and Technoloy Plasma Sources Sci. Technol. 24 (215) 222 (6pp) doi:1.188/ /24/2/222 Fast Track Communication Gravity effects on a lidin arc in four noble ases: from normal to hyperravity L Potočňáková 1, J Šperka 1,2, P Zikán 1, J J W A van Loon 3, J Beckers 4 and V Kudrle 1 1 Department of Physical Electronics, Masaryk University, Kotlářská 2, CZ-61137, Brno, Czech Republic 2 The Central European Institute of Technoloy (CEITEC), Masaryk University, Kamenice 5, CZ-625, Brno, Czech Republic 3 Dutch Experiment Support Center (DESC), ACTA-VU-University and University of Amsterdam, Amsterdam, The Netherlands and European Space Aency (ESA), ESTEC, TEC-MMG, Noordwijk, The Netherlands 4 Faculty of Applied Physics, Eindhoven University of Technoloy, P.O. Box 513, 56 MB Eindhoven, The Netherlands eorejewel@mail.com Received 2 October 214, revised 3 February 215 Accepted for publication 2 February 215 Published 19 March 215 Abstract A lidin arc in four noble ases (He, Ne, Ar, Kr) has been studied under previously unexplored conditions of varyin artificial ravity, from normal 1 ravity up to 18 hyperravity. Sinificant differences, mainly the visual thickness of the plasma channel, its maximum elonation and eneral sensitivity to hyperravity conditions, were observed between the dischares in individual ases, resultin from their different atomic weihts and related quantities, such as heat conductivity or ionisation potential. Generally, an increase of the artificial ravity level leads to a faster plasma channel movement thanks to stroner buoyant force and a decrease of maximum heiht reached by the channel due to more intense losses of heat and reactive species. In relation to this, an increase in current and a decrease in absorbed power was observed. Keywords: lidin arc, noble ases, hyperravity (Some fiures may appear in colour only in the online journal) Since plasma science mostly deals with chared particles, the ravitation force, which is much weaker than the electromanetic force, is often nelected. Althouh this miht be just and reasonable for many kinds of dischares, there are cases where ravity plays a crucial role throuh its action on neutral particles. Typical examples include astrophysical plasmas, dusty plasmas, arcs, plasma torches and others. In these dischares the alteration of real or apparent ravity (-force) could sinificantly chane the plasma properties with consequences for both fundamental and applied science. Study of the electric dischares in altered ravity could be important for electrical and fire safety precautions at hih accelerations (space fliht, aircraft, racin), ion thruster desin, or understandin the plasma related processes in the atmospheres of other planets, where the ravity is sinificantly stroner or weaker than that on Earth. One of the dischares where the ravity has to be taken into account is a lidin arc. A lidin arc is similar to well known stable arc dischare [1]. It differs in the electrodes eometry, which in the case of a lidin arc enables the vertical movement of the plasma channel. This movement is caused by: (i) ravity-dependent buoyant force oriinatin from the temperature difference between the hot plasma channel and cold surroundin atmosphere and/or (ii) the dra of as, which typically comes from below. In a standard lidin arc confiuration with diverent electrodes, the lenth of the dischare /15/222+6$ IOP Publishin Ltd Printed in the UK

2 channel increases as the arc is movin upwards. When the lenth exceeds a critical value, the dischare extinuishes and a new one is inited at the minimum distance between the electrodes at the same moment [2]. This repetitive character of a lidin arc lifecycle with a characteristic time period is one of its most distinuishin attributes. At low current density, i.e. in a so-called lowin arc [3] mode, lidin arc properties are similar [2] to non-equilibrium atmospheric pressure low dischares. The lidin arc has been studied intensively from the point of view of fundamental research [4 6], as well as an outstandin plasma source for various industrial applications [7, 8]. The lidin arc is usually studied in a flow reime with a relatively hih flow rate, which overns the dischare dynamics, especially the vertical speed of the plasma column [5, 9, 1]. In such a case, the buoyancy plays only a minor role and the as dra dominates. So to study the buoyancy effects, the ravity level should be increased to emphasise the buoyancy compared to the as flow effects. Several experiments have been performed to explore the ravity effects on standard arc dischares [11 13]. Only recently have the first studies [6, 14, 15] about the influence of increased ravity on the lidin arc behaviour been published. In hiher ravity levels the plasma channel lided faster and the whole life cycle was shorter. Here, an extension and improvement to our previous papers is made by carryin out the experiments in four noble ases He, Ne, Ar and Kr. These were chosen both for their similarities (they are all monoatomic and non-reactive ases) as well as for their differences (their atomic masses cover a wide rane, which results in a wide spread of dependent quantities, e.. thermal conductivity). The experiments in hyperravity (up to 18 ) were performed by usin the Lare Diameter Centrifue (LDC) at ESA-ESTEC centre in Noordwijk, the Netherlands [16]. The whole experimental setup (except the spectrometer) was placed inside the centrifue ondola. Given that the detailed description of the experimental apparatus and methods has been presented in [14], only a brief overview follows. The dischare at atmospheric pressure was operated between slanted copper electrodes with a minimum distance of 4.5 mm and an initial anle of 36 between them. The electrodes were enclosed in a non-conductive dischare chamber with 2 litre inner volume with holes for as inlet and outlet. The dischare voltae was supplied via a variable autotransformer ( 25 V, 5 Hz AC) by hih voltae transformer ( 1 kv). Three multimeters were used to measure the RMS voltae, current and power on the primary windin of the hih voltae transformer. A hih voltae probe and a Roowski coil current probe were used to measure the instantaneous electric parameters of the dischare. The dischare visual appearance was recorded as a video (up to 1 frames/s) and as standard photoraphs by a diital camera. Althouh the experimental device is essentially the same as in [14], the sealin of the dischare chamber was improved and the loner exhaust tube suppressed possible air backstreamin. Altoether, these chanes sinificantly improved the purity of noble as inside the dischare chamber. The amount of air admixtures in the noble as was estimated from spectroscopic measurement to be in the order of.1% or better. The main features of the lidin arc appearance will now be discussed based on fiure 1, where lon exposure (.1 1 s) photoraphs of the lidin arcs at various conditions in He, Ne, Ar and Kr are shown. Each row corresponds to one as and each column to different experimental conditions. The photoraphs in columns 1 4 were taken remotely from inside the centrifue ondola and they show the chanes that the lidin arc in each as underoes with the increase of artificial ravity. The experimental conditions were deliberately set differently for each as to enable lidin of the plasma channel already in 1 conditions and simultaneously not to mask the ravity effects by too hih a as flow. Unfortunately, due to severe complications (includin the influence of hyperravity on the camera lenses, shallow depth-of-field, impossibility to remotely control the camera durin experimental runs at LDC), the quality of some of the photoraphs was affected. The rihtmost column shows the photoraphs taken in laboratory 1 conditions with as flow increased by 1 slm compared to the columns 1 4 for each as. The chanes in brihtness between the imaes are not to be attributed to plasma processes but to different exposure times. In the majority of the photoraphs (except for Ar and Kr, both at 1 and 2 ), rather lon exposure time and hih lidin frequency caused an overlayin of many lides of the plasma channel. The most pronounced differences amon ases in the first column of fiure 1 are: apparent colour of the lidin arc plasma, which is naturally determined by dominant spectral lines of each as, the thickness of the plasma channel and its maximum elonation. The thickness of the plasma channel was observed to be visually wider for lihter ases (helium and ) and thinner for heavier ases ( and ). The phenomenon of radial contraction can be attributed to the inhomoeneous heatin in the radial direction due to finite thermal conductivity of the ases, which is more than 2 times lower for than for helium at 3 K and this difference increases even more at hiher temperatures [17]. As a result, the lidin arc in helium and had a rather diffusive character, while in and the sharp edes of individual filaments could be distinuished [15]. Unfortunately, quantitative analysis of the filament thickness is difficult: for the lon exposure photoraphs in fiure 1, the unceasin movement of the plasma channel causes motion blur of several millimeters and the low resolution of our hih speed camera leads to comparably hih uncertainty. The maximal elonation of the plasma channel before extinuishin was also different for each as. Generally, the lidin arc in was able to rise much hiher than the lidin arc in lihter ases. In typical conditions (low -level, the as flow low enouh to maintain reularly shaped individual plasma channels) the lidin arc in and was even able to easily ascend above the electrodes meanin that the end points of plasma channels reached the top of electrodes, yet the centre of the arc continued to rise up (see imaes in [15]). This can probably be attributed to the lower 2

3 Fiure 1. Lon exposure photoraphs of the lidin arc in helium,, and. Experimental conditions for each of the columns 1 4 (dependence on artificial ravity level) were set to enable lidin of the arc already in 1 : helium 1.37 slm, 8 kv;, and.4 slm, 4 kv. Conditions in column 5 differ in the as flow, which was increased by 1 slm for each as. Exposure times of photoraphs in columns 1 4 were: He.1 s, Ne 1 s, Ar.1 s, Kr.3 s; exposure times of photoraphs in column 5 were: He.1 s, Ne.1 s, Ar.1 s, Kr 1 s. ionisation potential of heavier ases, which makes sustainin the dischare easier than in helium or. The influence of increasin hyperravity can be observed on photoraphs in columns 2 4 of fiure 1. With the increase of ravity, the ravity-dependent buoyant force actin on the plasma channel ets stroner, makin its upward movement faster. However, due to faster convective coolin, the losses of excited atoms and heat become more intense and cause the decrease in maximum heiht. This decrease is the most sinificant for and the least sinificant for helium. This different sensitivity of each as to an increase in ravity is the result of various factors, one of them bein the as atomic weiht. The resultin ravity-dependent force (the difference between downward ravitational force and upward buoyant force) actin on the heated arc channel is proportionally dependent to the mass density difference between a cold and a heated noble as. This difference is proportional to the atomic weiht of the as (4, 2, 4 and 84 for He, Ne, Ar and Kr) and so the hyperravity effects are much more profound for than for lihter ases. Also, the thin filament of transfers the heat to the colder environment more slowly than the wide diffusive helium plasma channel. This effect is intensified by the fact that helium is a ood heat conductor compared to. Thus, the temperature difference between the heated and cold as is bier for heavier noble ases, therefore enhancin the hyperravity effects. Finally, the thinner the plasma filament, the less as aerodynamic resistance it has to overcome durin its upward movement. In the lidin arc device with the as feed from below, the plasma channel is lifted up by a combination of two forces: buoyant force and as dra. Increasin either of them will make the resultin upward force actin on the 3

4 maximum heiht [cm] , normal as flow 18, normal as flow 1, hih as flow He Ne Ar Kr Fiure 2. The maximum heihts of lidin arcs calculated from imae analysis of fiure 1 in 1, 18 and hih as flow conditions. plasma channel become stroner. Which component of the force will dominate depends on the -level and on the as flow rate. The hih ravity case was already discussed and the hih-flow case (approximately 1 slm above the normal flow conditions presented in columns 1 4 of fiure 1: helium 1.37 slm 2.37 slm;, and.4 slm 1.4 slm) is presented in column 5 of fiure 1. The hih as flow influences the motional characteristics of the lidin arc as the plasma channel is draed by the fast movin as. It was found that even thouh this force is of a different nature to the buoyant force the main effect on the lidin arc remains the same the speed of the plasma channel increases, convective coolin also increases and as a consequence the maximum heiht of the arc decreases and new initions occur more frequently. The recapitulation of maximum heihts in four ases in 1 versus 18 versus hih as flow from fiure 1 can be seen in fiure 2. As discussed above, the increase of the artificial ravity level did not cause similar chanes in all four ases. The decrease of maximum heiht in helium and is only small (around 1%), while in and the decrease is sinificant (over 5%). Krypton is the only as in this experiment, for which the increase of artificial ravity caused a bier decrease of the maximum heiht than did the increase of the as flow. In the above-described experiment, the experimental conditions for individual ases differed in order to maintain the lidin motion of the plasma channel. If the conditions were set the same for each as, then the situation miht be drastically chaned, as will be shown in fiure 3. Here, the as flow rate was.15 slm for each as and the dischare voltae, 4 kv, was also the same for each as. Besides the maximum heiht, the frequency of new initions (or lidin frequency) (fiure 3(a)) is also a valuable parameter for the study of hyperravity influence because it effectively combines both the maximum heiht reached by the plasma channel (fiure 3(b)) and the velocity of its movement (fiure 3(c)). For these particular experimental conditions, the lihtest of the ases helium appeared in the form of stable arc-like plasma channel for all conditions (zero lidin frequency) because the upward force was not stron enouh to lift the arc up. The second lihtest as also started as a stable plasma channel but was able to undero a transformation to a lidin arc after an increase of the artificial ravity level above 8. The buoyant force for the dischare in and was stron enouh to initiate lidin already at 1, althouh the frequency was quite low, since the plasma channel moved rather slowly and quenched (and reinited) after travellin a relatively lon distance. With increasin ravity, the lidin frequency increased to values many times hiher for all three ases for which the dischare appeared in the form of a lidin arc. However, one must keep in mind that the lidin frequency is a time-averaed quantity, so it exhibits a statistic behaviour and, hence, the individual lides may differ a lot. Moreover, the lidin arc is also influenced by the system history (e.. preheated electrodes, as circulation) and shows sins of hysteresis. We repetitively observed that the transition from stationary into lidin mode happens at a hiher as flow rate than reverse transition (from lidin into stationary mode), probably thanks to the positive feedback between the ambient as circulation and plasma channel lidin. The electrical parameters that were directly measured on the primary windin of hih voltae transformer are plotted in fiure 3(d). Since the dischare in helium was not in a form of a lidin arc but a stationary arc-like dischare, it is not included in the raphs, the same oes for in -levels below 8. The curves for and are almost constant, while for they exhibit a clear dependence. In, as the -level increases, the power input into the primary windin (see fiure 3(d) solid symbols) decreases as a consequence of the lowered power consumption in the ravity-shortened plasma channel [2]. Simultaneously, the primary windin current increases (see fiure 3(d) open symbols), albeit slihtly. This counter-intuitive behaviour is most probably caused by the non-ideality of the transformer, i.e. its transformation matrix is not diaonal, it is complex and may even be non-linear. The substantial primary current error bars at low -levels would then be a consequence of the sensitivity of the current to the phase variations for a iven voltae and power at low power factors. The plots of primary current and RMS power for and are almost constant with -level because their heiht decreased insinificantly. In this paper, we have shown that the increased artificial ravity (up to 18 ) spectacularly influences the properties of the lidin arc dischare between diverin electrodes. Four noble ases of atomic weiht in wide rane (He, Ne, Ar and Kr) were used to reveal how the differences in lidin arcs observable at 1 affected their peculiar behaviour at hiher levels. At the elementary level of the complex interplay of physical phenomena evoked by hyperravity, there is an increase of the ravity-dependent buoyant force. The stroner buoyancy initiates faster movement of the plasma channel, which 4

5 lidin frequency [Hz] helium maximum heiht [cm] helium (a) (b) plasma channel velocity [cm/s] (c) helium primary power [W] (d) power: current: 2, ,4 2,3 2,2 2,1 primary current [A] Fiure 3. (a) The lidin frequency, (b) maximum heiht reached, (c) velocity of the plasma channel and (d) primary windin power and primary windin current of the lidin arc in He, Ne, Ar and Kr at low as flow (15 sccm) and 4 kv dischare voltae, as a function of the ravity level. increases the losses of excited species and heat and thus results in the decrease in maximum heiht reachable by the plasma channel. These chanes of the plasma properties are then reflected in the electrical parameters of the dischare. All of these effects of the varyin ravity on the lidin arc dischare are stronly interconnected and create a complex system that is very sensitive to all of the experimental conditions, even with sins of hysteresis behaviour. Acknowledments This work was supported by European Space Aency SpinYourThesis! 213 proramme, Masaryk University Rector s Scholarship, project CZ.1.5/2.1./3.86 funded by European Reional Development Fund and project LO1411 (NPU I) funded by Ministry of Education Youth and Sports of Czech Republic. We would also like to thank Mr A Dowson from ESA-TEC-MMG for his support of these studies. References [1] Raizer Y P 1991 Gas Dischare Physics (Berlin: Spriner) [2] Fridman A, Nester S, Kennedy L A, Saveliev A and Mutaf-Yardimci O 1998 Glidin arc as dischare Pro. Enery Combust. Sci [3] Indarto A, Choi J-W, Lee H and Son H K 26 Effect of additive ases on methane conversion usin lidin arc dischare Enery [4] Richard F, Cormier J M, Pellerin S and Chapelle J 1996 Physical study of a lidin arc dischare J. Appl. Phys [5] Zhu J, Sun Z, Li Z, Ehn A, Aldén M, Salewski M, Leipold F and Kusano Y 214 Dynamics, oh distributions and uv emission of a lidin arc at various flow-rates investiated by optical measurements J. Phys. D: Appl. Phys [6] Šperka J, Souček P, Van Loon J J W A, Dowson A, Schwarz C, Krause J, Butenko Y, Kroesen G and Kudrle V 214 Hyperravity synthesis of raphitic carbon nanomaterial in lide arc plasma Mater. Res. Bull [7] Czernichowski A 1994 Glidin arc: applications to enineerin and environment control Pure Appl. Chem [8] Kalra C S, Gutsol A F and Fridman A A 25 Glidin arc dischares as a source of intermediate plasma for methane partial oxidation IEEE Trans. Plasma Sci [9] Bo Z, Yan J H, Li X D, Chi Y, Chéron B and Cen K F 27 The dependence of lidin arc as dischare characteristics on reactor eometrical confiuration Plasma Chem. Plasma Proc [1] Sun Z W, Zhu J J, Li Z S, Aldén M, Leipold F, Salewski M and Kusano Y 213 Optical dianostics of a lidin arc Opt. Exp [11] Suits C G and Poritsky H 1939 Application of heat transfer data to arc characteristics Phys. Rev

6 [12] Flikweert A J, Nimalasuriya T, Kroesen G M W, Haverla M and Stoffels W W 29 The metal-halide lamp under varyin ravity conditions measured by emission and laser absorption spectroscopy Microrav. Sci. Technol [13] Tan G D and Mieno T 21 Synthesis of sinle-walled carbon nanotubes by arc-vaporization under hih ravity condition Thin Solid Films [14] Šperka J, Souček P, Van Loon J J W A, Dowson A, Schwarz C, Krause J, Kroesen G and Kudrle V 213 Hyperravity effects on lide arc plasma Eur. Phys. J. D [15] Potočňáková L, Šperka J, Zikán P, van Loon J J W A, Beckers J and Kudrle V 214 Glidin arc in noble ases under normal and hyperravity conditions IEEE Trans. Plasma Sci [16] van Loon J J W A, Krause J, Cunha H, Goncalves J, Almeida H and Schiller P 28 The lare diameter centrifue for life and physical sciences and technoloy Proc. Of the Life in Space for Life on Earth Symp. (Aners, France, June 28) ESA SP-663 [17] Kabouzi Y, Calzada M D, Moisan M, Tran K C and Trassy C 22 Radial contraction of microwave-sustained plasma columns at atmospheric pressure J. Appl. Phys