Experiments on Gas Jet in the Wendelstein 7-AS Stellarator

Plasma Science Technology, Vol 5, No 5(2003)Experiments on Gas Jet in the Wendelstein 7-as Stellarator*Yao lianghua(姚良骅)1, J. Baldzuhn21 Southwestern Institute of Physics, Chengdu 610041, China0+ AMax-Planck Institut fuer Plasma physik, EURATOM Ass, 17491 GreifswaldGermanyAbstract Wendelstein 7-AS(W7-AS)pertains to an advanced helical stellarator. A new fu-elling method, the supersonic molecular beam injection(SMBI, named Gas Jet in Germany)system was installed in W7-AS in May 2001 as a cooperation research item co-supported by theNational Nature Science Foundation of China and the Max -Planck Institute of Plasma PhysicsGarching, Germany. The experiments of the gas jet with hydrogen or deuterium on W7-AS wereimplemented. The experimental results exhibit the following features such as high fuelling ef-ficiency, stable high-density plasmas and reduction of the recycling fluxes from the vessel wallduring injection. These crucial points show that the new fuelling method can be applied to longand stable discharges.Keywords: gas jet(Supersonic Molecular Beam) injection, wendelstein 7-AS stellarator,fuellingPACs:52.50G,52.55.H1 Introductioncooperation research co-supported by the NationalNature Science foundation of china and the maxOn the 19th IAEA Fusion Energy ConferencePlanck Institute of Plasma Physics, Garching, Ger(Lyon, France, 14 N 19 October 2002), Prof. Fmany, a new fuelling method, the supersonic molecWagner said(l:"After nearly 14 years with 56953gas jet indischarges, W7-AS has been suspended from opera- many)systern 2) had been applied to W7-AS in Maytion.The device is now mothballed. The develop- 2001. The experiments of the gas jet with hydrogenment of the Wendelstein stellarator line will go on or deuterium on W7-AS have been implemented forwith the w7-X device. The major goals of W7-ASare to test the modular coil concept, to demonstratethe effectiveness of the first steps toward an opti- 2 Magnetic confinement andmized stellarator design, and to develop an exhaustheating system of W7-ASconcept based on the nat ural island chain which firmsup the plasma boundary. W7-AS has contributed toW7-As pertains to an advanced helical stellaratorareasof stellarator and fusion researchwith a major radius of 2 m and a plasma radius inTwo years before the W7-AS was mothballed. a the range of 0.11 m to 0. 18The project supported by the National Nature Science Foune中国煤化工1005060 and the ChinaNuclear Industry Science Foundation( No. 94C03033)YHCNMHGYao Lianghua et al.: Experiments on Gas Jet in the Wendelstein 7-AS Stellaratormoula" corner coiltriangularplatelliot cal planeLasma tr Midal lichI ccilFig 1 Magnetic confinement coil system and limiters of W7-ASstructure of w7-As with limiters is shown as inThe orientation of gas jet injection(GJI)systemFig 1. W7-AS is a flexible system comprising 45 is on the sectorof M5 with a toroidal angle of-254omodular cojich produce both toroidal and and a poloidal angle of-28.poloidal field components with [=2 and l=3 injection system. The valve used for producing gasdominant poloidal components. An additional sejet is a fast gas valve(type series 99, General Valveof toroidal field coils enables change of rotational Corp, Fairfield, USA)with a Laval nozzle orificetransform. Five separately operated modular coilsof 0. 1 mm diameter, which is installed on a rect-at the corners of the pentagon-like shaped system angular port of the vacuum vessel, as shown in Figwith a n=5 toroidal symmetry allow variation of 2. No skimmer is used in front of the orifice. Thethe toroidal mirror ratio. The plasma position can distance between the nozzle and the edge plasma isbe changed with vertical field coils, and an ohmic about 108 cm, which is two and a half times longersystem was used to modify rotational transform, to than the distance for the beam injection in HL-1Mstudy current-driven instabilities, or to compensatetokamak. To maintain a low background pressurethe bootstrap current. Five pairs of in-vessel controljet injection, a turbo-molecular pumpcoil(top-bottom) being allowed to change the Bspacity 31/sec) is installed half-way be-field spectrum, were used to vary the edge island size tween the plasma and the fast valve. To protect thealong with its connection length employed for diver- fast valve against stray magnetic fields from W7-AStor operationcoil system, a 5 mm thick shield of tempered steelPlasma heating of W7-AS was done with ECRHype ST33)is applied around it. Beside the fast(70 GHz, 140 GHz, 2.4 MW absorbed power, NBI valve, a standard video camera(type Panasonic GP2.8 MW absorbed power, in the last phase with KS162HDE)is installed on the common flange withall-co or all-ctr orientation), and ICRH(up to 1 Mw the fast valve, the camera looks in parallel to theheating conditions.For the experiments the gas pressures between 0.5MPa and 5 MPa are applied and the working gas is3 Arrangement of gas jet injechydrogen or deuterium. The pulse duration of thetion system in W7-AS中国煤化工nd 40 ms. A system ofors IsCNMHG1934Plasma Science Technology, Vol5, No 5(2003)W7 AS AdvanTharrosCXRTO GHz ECRHSof-x Rosendal angia -cW7-AS plasmaMach shock froneNone of silenceNozzle fast valveFig 2 Heating and diagnostics system of W7-AS as well as the poloidal cross section of W7-AS and GJI systemdistributed around W7-AS. The nearest one of themDischarges with pure ECR heating are performedis located about 50 cm away from the beam injection to compare the property of the gas jet feeding withort, looking on the inner vessel wallthe standard gas puff. The discharges are startedwith a standard gas puff, which is switched off fo4 Experimental result150 ms to feed the plasma exclusively with GJI. InThe video frames show a well-defined bright areapuff is switched on again. During the discharges, thewith a diameter of a 6 cm during GJI on the plasma profiles of electrosurface, indicating a good location of the injected measured by Thomson scattering and ECE receivermaterial within an injection angle of 3 degreesThe electron line density is monitored中国煤化工CNMHGYao Lianghua et al.: Experiments on Gas Jet in the Wendelstein 7-AS StellaratorW7 AS shot55106(20024121237)500300100 Radiation loss powerTe edgeGas puff rate100200300400500600002003Fig 3 Time traces for shot 55106. The standard gas puff is switched off during the SMB injection, which is indicatedby the dashed line at t=250 ms. The valve opening time is 12 ms, The bolometric radiation power losses are notinfluenced by the gas jet. The Balmer-a light intensity is recorded in vicinity to the gas jet locationby HCN interferometer, and the ion temperatureslightly. Table l gives the density peaking factorsare measured by spectrometer. Experimental mea-versus time for shot 55106. The sampling time pointssurement on a series of identical plasma parameters are determined by the YAG Thomson system. Theare performed, where the operation time of the fast relative changes of P are given with respect to the Pvalve is varied between 6 ms and 18 ms. The hy-value at t= 230 ms before the beam injectiondrogen pressure is kept constant at 3.0 MPa. TheThe maximum peaking factor P depends stronglytime traces of such a discharge are shown in Fig 3. on the fast-valve opening time. The longer the fastWhen the supersonic beam is injected at 250 ms im- valve is open, typically the higher the obtained peak-ediately after injection, the density and diamag-ng factor is. The energy confinement time TE for thenetic energy tend to increase further during a shortdischarges is, however, not substantially improvedtime interval of a few ms. Accordingly, Te decreases Before the beam injection, the discharges withinstronger in the center than that at the plasma edgethe measured error bars obey the modified Lackner-The strong peak in the Ho trace indicates the lo- Gottardi scaling law. After the injection at the mo-al external particle source. The profile peaking fac- ment of maximum diamagnetic energy, in some casestor defined by P=no(O)/ne is calculated from the the measured TE is up to 10% higher than the scalingmeasured ne profile data each 50 ms by the YAclaw. However, considering the measured error barsThomson scattering. Typically, the profiles are even this improvement is still smaller than the margins offattened by N 10% directly after the gas jet injection, but after 50 ms or later they tend to peak中国煤化工CNMHG1936lasma Science Technology, Vol 5, No 5(2003)2 1Change energy confinement time /%Duration gasjet opening time /msFig4 Linear fits to the fuelling efficiency and the change of the density peaking factorTable 1. Relative changes of the density peaking clear orifice the supersonic jet can be established infactor vs time for shot 55106a perfect manner. Though both effects are small, theTime Peaking factor Relative changesolid lines fitting linearly with the data shows thatPof P(%One crucial point is the behaviour of the recy-cling fuxes from the vessel wall during the jet in-tion. Fig. 5 shows time traces of the ho detec-1.51tors, of which one is adjacent to the gas jet(upper1.266left), and the three others observe the vessel wall ata distance. Directly after the injection, the H, in-The upper plot in Fig. 4 shows the relative changeensity drops at a distance from the beam injectionof the density peaking factor normalised to l versus location and is recovered after N 150 ms to the valuethe relatively enhanced TE in % The stronger the before the injection. The T. and Ti on the last closeddensity peaking after injection is the higher the im- Aux surface, however, are reduced only slightly by Nprovement in TE will be showing the relation between20% during an initial injection period of a 20 ms,asthe density peaking and the improvement of the en- well as ne close to the plasma edge.Therefore, theergy confinement. The dependence of the fuelling ef. temperatures are still high enough to allow for theficiency on the valve opening time, which is roughly interpretation of the Ha time traces as local particleproportional to the amount of material provided, are fluxes. This indicates a transient reduction of theshown in lower plot of Fig. 4. The longer the openecycling Huxes during the beam injection. It is sup-ing time, the higher the efficiency becomes, possibly ported by neutral gas manometer measurements, byup to 23%. By way of explanation, it is assumed thatwhich a reduction of the neutral pressure after thethe gas jet requires a few ms to be established in fullbeam injection is registered. This could point to aperformance. In particular, the fast valve will need atransient relief from the wall recycling in favour offew ms to open completely. Only with a completelyhy the fIr中国煤化工CNMHG1937Yao Lianghua et al.: Experiments on Gas Jet in the Wendelstein 7-AS StellaratorW7 AS shot55106(20024121237州小W叫t/mst/Fig 5 Ha time traces, gas jet injection time is at t= 250 ms(dashed line)5 Summaryder construction, but the injection system must befurther modified including a density feedback controlThe principal results of GJI experiment in W7-AS systemarea. This is a useful fuelling method of obtaininghigh fuelling efficiency and stable high density plaReferencesmas for WZ-ASb. The crucial point is the behaviour of the re-1 Wagner F, et al. Major Results from Wendelsteinycling fluxes from the vessel wall during the beam7-AS Stellarator, in 19 th IAEA Fusion Energyinjection, with which a reduction of Hu intensity andConference, Lyon, France, 14 N 19 October 2002neutral pressure shows that the new fuelling methodAEA-CN-94/OV/2-4can be applied to stable discharges of long duration2 Yao L H, et al. Nucl. Fusion, 2001, 41: 817c. GJI may be considered as an available fuelling (Manuscript received 13 May 2003)ethod for the large stellarator W7-x, which is un- E-mail address of Yao Lianghua: yaolh@swipaccn中国煤化工1938CNMHG