ABSTRACT. 1 Introduction for w < w and is classified as type II. Type II discharge is

Transcription

1 An ICRF Plasma Heating Experiment in the Tandem Mirror Propulsion Device T. F. Yang. S. Peng, F. R. Chang-Diaz,* MIT Plasma Fusion Center, Cambridge, MA 02139, USA *Astronaut Office, NASA, Lyndon B. Johnson Space Center, Houston, Texas 77058, USA ABSTRACT the type 11 discharge while exhibiting modulated amplitude. Similar phenomena were observed through the electron den- sity measurement using a microwave interferometer. Recently a laser fluorescence system for neutral density measurement has become operational. A test measurement of the electron temperature, density and neutral density pro- file has been made. This is a very important step for future exhaust study. ICRF plasma heating experiments have been carried out on the tandem mirror plasma propulsion device at MIT. The measured radial profile of the wave amplitude agrees with theoretical predictions. The axial profile was measured by an array of B-dot probes mounted in the region where the damping would occur from the theoretical prediction. The measured wave was found indeed damped when it approached the resonance plane indicating the beach heating effect. This ef- feet is again confirmed by an analytical solution of the wave on a slab model. The time characteristics of electron density, measured by a microwave interferometer, confirm the two types of discharges observed from H. emission reported 2 ICRF Wave Experiment previously. The plasma is more stable and has higher density for the type 11 discharge. From these two results the type II The experimental setup is shown figure 1. The wave is discharge is chosen discharge a as the is preferred chosen plasma heating method. launched from the antenna in the high field region and prop- The stability phenomena have also been studied by varying agates toward the resonance plane and inward toward the the magnetic field intensity and was found that the plasma axis. The measurements were made at the resonance plane was stable at higher field. A preliminary study of the plasma where the diagnostics are currently located. in the exhaust has been made. The eletron temperature is The electron density of the plasma at two different ICRF 40 ev ( z 463,768 K ) in the exhaust. Both electron density frequencies was measured with a microwave interferometer and temperature radial profile are nearly flat. The neutral and is shown in figure 2. The characteristics of the density density, measured by laser fluorescence method, peaked near results reconfirms the phenomena of the existence of two the edge. types of discharges observed from the H. emission reported previously [8]. Figure 2a shows the discharge for w > w, and is classified as type I, whereas figure 2b shows the discharge 1 Introduction for w < w and is classified as type II. Type II discharge is more quiescent than type I. The production and control of the necessary plasma param- Type I discharges generate densities that are lower than eters to generate variable I,, and thrust at high power is type II discharges at 50 kw of ICRF heating. The peak line the fundamental rational for the development of the ICRF density obtained from the interferometer is approximately heated Tandem Mirror Plasma Rocket. To this general end, x 10 1 cm-', which corresponds to a plasma density an experimental program has been underway at MIT since f approximately 1.6 ± 0.42 x 10"cm-. The average line 1988 [1],[2],[3],[4],[5]. These experiments have shed new light density is approximately x 10"1cm - or x on the dynamics of these high-power devices and their ap- 10"cm 3. The uncertainty is due to both pickup as well as plication to multimegawatt space propulsion. the resolution of the phase digitizer. Much higher densities An important finding leading to a potential enhancement would of course be obtained as the power level is increased. of the previously expected ICRF efficiency is the so-called The radial profile of B, of the wave for the the type II "BEACH" effect[6]. This phenomenon predicts significant discharge measured by a single B-dot probe at midplane is rf wave absorption by the plasma while enroute to the ion shown in figure 3. The solid curve shows the numerical recyclotron resonance plane. Hence, the optimum antenna lo- sults and the open circles with crosses are the experimental cation is not at the resonance planes as predicted earlier, but data. They are in good agreement. in fact somewhere near the high field region. This theoretical The axial profile of B. of the ICRH wave for the type prediction reported earlier has been recently confirmed ex- discharge measured by the B-dot probe array is shown in perimentally by an assortment of new diagnostics including figure 4. It is seen that the wave is damped as it travels toan axial array of 5 "B - dot" probes built and installed this ward the resonance plane. The numerical simulation of the year. wave was presented previously [7!, [8j and a trimetric view It was also reported previously [7],[81that there were two of the calculated B, is shown in figure 5. The corresponding types (labeled I and II) of discharges observed from the mea- region of the axial profile measured by the B-dot probe arsured H. emission. There was fluctuation during the initial ray is indicated in this figure as the damped region between phase of the type I discharge. The fluctuations decayed and the peak and resonance plane. The wave is rising to a peak the plasma became quiescent. This was an indication of and is then damped toward the resonance plane in agreeinstability at the beginning of the discharge. but was stabi- ment with the experimental data. Such a damping effect is lized after. On the other hand, the plasma was quiescent for also predicted by the analytic solution discussed in the next sect ion. 1

2 Laser Floreseiicc ICIIII A unnuja ECuII Lan I mirprotrmiple' Probe Gas U-dot Probe Mixcrowave Interferonieter!.angintiir Probe Figure 1: Layout of the ICRF wave experiment u.6 N-i mode. )HLP Antenna. rype ii x Zv 0., ID~ THEORY TME(ms) L Figre ensty : Pa.~a easredbyintrfeomeerforfigre : Rdil B prfil fo tpe I dschrg (atp3. icare(83ean b yei dshre( E 0.50 XPERMEN

3 The analysis of wave propagation in the slab geometry is much simpler than that in the cylindrical geometry and can ' provide information on the coupling efficiency of ICRF power Sto plasma. The detailed analysis will be given in reference 61.. :.0 - A figure of merit was defined as the dissipated power divided -by the source power as given by the incident Poynting Flux. Since for th majority of the cases. ; I, > v,. the much 7 ] simplified 7 is obtained o - 1 : =*' rf - 2 ' 9 T ) r J (r0s9; -" cost) 2 " - j sin'9,. ".(1) o.o L * '6 50 where OfT are the incident and transmitted angles of the 1 power flux and are defined as i st. rcm ont. (cm' 1 os' 0, -- cos '( ) ( Figure 4: Axial B, profile for type II discharge where S2 D2 S 2 _D - SS \Ve will examine the following two important limits: 1. Resonance limit (w f l,) In this limit, it is seen that S, D - o. This allows the dropping of the last term in the denominator. It is - ^ also seen that cosotr I, and k >> ko. This results in the 2 Y SS = sk- (2) Since S. ~- 1, DR, i 0 in the limit of resonance, it can * ''f be shown that the efficiency T approaches a maximum of 1 when resonance is approached. '" 1'C 2. Far from the Resonance Limit (w > f, w < Qi) // / _ For frequencies far away from resonance, the value of -* the efficiency can be given by the same equation as re.,, " before (Equation 2). However, the simplification of S-D (3) ' can still be made. This results in an efficiency of F - Q, cos OT (4) Figure 5: Trimetric view of B, from numerical simulation. --P \/i The heavy solid curve corresponds to the B-dot probe array Since for our range of frequencies, 0, w«, for fremeasurement quencies off resonance, the efficiency approaches zero. 3 Theoretical Analysis The results obtained for the slab geometry show that F increases as resonance is approached. The effort of the theoretical work is to obtain much more rigorous treatment of the ICRF wave propagation than previously reported. The true geometry can be considered in In other words. large wave power is dissipated or the wave is damped. These theoretical predictions agree with preliminary experimental results of the axial B. profile. numerical simulations, but the physical insight is difficult to extract. The functional dependence of the wave dissipation on the plasma parameters can be formulated analytically only after many simplified assumptions are made. Notwith- 4 Preliminary Study of The plasma standing these simplifications building upon the relatively crude analytical solution of slab and cylindrical models which exhaust was presented previously [9],[8. The refined analysis of both models was undertaken and te results of the slab model i [The laser fluorescence method for measuring the neutral denmodels was undertaken and the results of the slab model is i r T. sity has been used in the Tara Tandem Mirror fusion energy discussed in the following. 3

4 experiment [10. The same laser system has been recondi- 5 Conclusion tioned and set up in the exhaust end of this device(figure 1). The fluorscence signal is very weak. To reduce the interfer- From the discussion given above an efficient method of heatence from the scattered light a two path optical method has ing the plasma in the tandem mirror plasma propulsion debeen developed. A clean fluorescence signal can obtained by vice can be developed. A wide range of plasma parameters taking the difference of the signal from the two optical paths, will be obtained in order to achieve a wide range of specific This improvement allows us to make successful measurement impulse and thrust propulsion [12]. The fuel injection and of the neutral density in the exhaust [11]. The fluorescence exhaust study will be carried out when the projected plasma signal detector can be scanned radially. The electron tem- parameters are established. perature and density were also measured with a Langmuir probe at the same radial position as the fluorescence detector. The test results are presented in figure 6. Presently the fuel is injected radially and axial fuel injection Acknowledgment has not yet been built and the exhaust has not been This work was jointly supported by the Office of Scientific connected to the large vacuum tank. The plasma is not flow- Research, US Air Force, and The US National Aeronautics ing. The tandem mirror is operated at a mirror ratio of 10, and Space Administration through Jet Propulsion Laboratherefore the plasma is well confined. The exhaust plasma tory and Johnson Space Center. is a result of loss cone effect. Thefore the density in the exhaust is only 10% of the central cell. However, the electron temperature is almost the same as that in the central cell. Both the density and temperature are almost flat as function References of radius which indicates that there is no loss of efficiency due to profile effect. The neutral radial profile is hollow, i.e., [1] F.R. Chang and J.L. Fisher. Nuclear Fusion, 22(8), it peaks near the edge and is nearly zero on the axis. This phenomena agrees with our prediction that the wall is insu- [2] F.R. Chang, W.A. Krueger, and T.F. Yang. In lated by neutrals. The density in the exhaust will be as high AIAA/DBLR/JSASS Int. Electric Propulsion Conferas the central cell when the device is made to operated at ence, paper AIAA , Alexandria, streaming plasma mode. More detailed study of the exhaust properties, such as the velocity, thrust and mass flow rate, [3] T.F. Yang, R.H. Miller, K.W. Wenzel, and W.A. is underway. Krueger. In AIAA/DBLR/JSASS Int. Electric Propulsion Conference, paper AIAA , Alexandria, [4] F. R. Chang-Diaz, T. F. Yang, W. A. Krueger, o. S S. Peng, J. Urhahn, X. Yao, and D. Griffin. In DGLR/ )o AIAA/JSASS Int. Electric Prpulsion Conference, pa- S (a) per DGLRA , Garmisch-Partenkirchen, W. Ger- 0 many, R(cm) [5] T.F. Yang, F.R. Chang-Diaz, S. Peng, J. Urbahn, _ and X. Yao. In AIAA/ASME/SAE/ASEE, 25th Joint 3 T - Propulsion Conf., paper AIAA , Monterey, CA, ST T i S () [6] Scott Peng. Wave propagation in tandem mirror plasma S propulsion device. Technical Report To be Published, Plasma Fusion Center, z o... I......,,.., [7] F.R. Chang-Diaz and T. F. Yang. Final technical re- R(m 2 port on propulsion research on the hybrid plume plasma rocket. Technical Report to be published, Plasma Fu- 'sion Center, S (c) -_ [- [8] S. Peng T.F. Yang and F.R. Chang-Diaz. The propa- 1 gation of rf wave in a tandem mirror plasma propulsion 6 - device. In AL4A/DGLR/JSASS, 21th Joint Propulsion o / Conf., paper AA.4A , Orlando, Fl, /[9] F.R. Chang-Diaz and T. F. Yang. Final technical re- - // port on propulsion research on the hybrid plume plasma -/ rocket. Technical Report PFC/RR-89-15, Plasma Fu- z I,.sion 3 Center, R(cm) [10] W.C. Guss, X.Z. Yao, L. Ptcs, R. Mahon, J. Casey. and Figure 6: The electron temperature T., density n., and neu- R.S. Post. Rev. Sci. Instrum., 59:1470, tral density n, in the exhaust chamber 4

5 [11. X.Z. Yao. T.F. Yang, and F.R. (hang-diaz. Neutral hydrogen density measurement in tnispp. Technical Report PFC/JA-91-23, 'MIT, I2 F.R. Chang-Diaz and T.F. Yang. In To bt prrstnted at.4 ID.4AIA A/DGLR/SASS 22nd Int. El~tn-c PTopulsion Conjtrence, Viareggie, Italyv