High resolution temperature and density proﬁles during the … · 2020. 7. 29. · (3mm), Thomson scattering proﬁles obtained during the energy quench of fast precursor disruptions.

PHYSICS OF PLASMAS VOLUME 9, NUMBER 8 AUGUST 2002

High resolution temperature and density profiles during the energy quenchof density limit disruptions in Rijnhuizen tokamak project

F. Salzedas,a) S. Hokin,b) F. C. Schuller, A. A. M. Oomens, and the RTP TeamFOM-Instituut voor Plasmafysica Rijnhuizen, Association Euratom-FOM, Trilateral Euregio Cluster,Nieuwegein, The Netherlands

~Received 18 January 2002; accepted 1 May 2002!

Measurements of the electron temperature,Te , and density,ne , during the energy quench of a majordisruption showed that the onset ofTe erosion in the neighborhood of them/n52/1 O point at thelow field side~LFS! accelerates the well-knownm/n51/1 erosion of the core temperature. Duringthis phaseTe(r ) is only partially flat in the region between theq52 and theq51 surfaces andne(r ) decreases in the core and increases inside them/n52/1 island. Immediately after theflattening ofTe(r ) a large peak inTe and to a lesser extent inne has been observed. This peak isradially localized at theq52 radius at the LFS, is very short lived and is poloidally asymmetric.Te

profiles measured by the heterodyne radiometer and the Thomson scattering agree very well up tothe timeTe(r ) flattens but afterwards can be a factor of two different. ©2002 American Instituteof Physics. @DOI: 10.1063/1.1489675#

itha

tolle

ao

todathreve

ta

o

a

im.

xp

t atise

re

ofncemege.m-w

meisde-nchn-sor,thee-e

file

isrt-

arein

s-

gy

tailthe

ay,

-

ion

I. INTRODUCTION

The power yield of a tokamak reactor increases wplasma density. However, since the beginning of tokamresearch, it has been found1 that high density plasmas tendbe unstable and can be abruptly lost during a so-caplasma major disruption. In the small tokamaks availablethe time, disruptions only resulted in an early terminationthe discharge. As the size of the tokamaks and also theirplasma energy increased it started to be clear that majorruptions are a threat to the structural integrity of a tokamreactor. In today’s tokamak operation a careful choice ofplasma parameters has decreased significantly the occurof disruptions and the related damaging effects. Howeevery disruption is undesirable and the goal is to decreaseprobability of occurrence of disruptions as close to zeropossible. So the study of major disruptions is of great imptance for the future of tokamak reactors.

Disruptions are thought to occur due to the nonlineinteraction of MHD ~magnetohydrodynamic! instabilities.Nonetheless, a detailed physical understanding of their tevolution has still not been fully accomplished2

Experimentally,3–11 the difficulties in probing the disruptionare mainly connected with its extremely short duration~mi-croseconds!, unpredictable onset and 3D structure. Also, ecept for electron temperature and density, local plasmarameters are difficult to measure, or are not measured aTheoretically,12–17the description of the disruption is still nounique. Different models address different parts of the druption but there is no continuous description of its tim

a!Present address: Centro de Fusa˜o Nuclear, Association Euratom-IST, Lisbon, Portugal; electronic mail: [email protected]

b!Alfven Laboratory Division of Fusion Plasma Physics, AssociatEuratom-NFR, Royal Institute of Technology, Stockholm, Sweden.

3401070-664X/2002/9(8)/3402/11/$19.00

Downloaded 31 Jul 2002 to 193.136.136.136. Redistribution subject to A

k

dtftalis-kencer,hes

r-

r

e

-a-ll.

-

evolution. These difficulties just reflect the complex natuof the phenomenon.

Density limit disruptions are preceded by a sequenceevents, called the precursor of the disruption. This sequeis not exactly the same in all tokamaks or even in the satokamak from disruptive discharge to disruptive discharDespite the large difference in sizes and the different geoetries in today’s tokamaks, density limit disruptions shocommon features. Typically anm/n52/1 tearing mode isdestabilized by the contraction of the current profile. At soinstant during the mode evolution when its amplitudelarge, the energy confinement of the plasma is abruptlystroyed. This process is usually called the energy que~EQ!. So far observations show that it is irreversible, cotrary to the course of events that occur during the precurwhich can be avoided. The EQ signals the beginning ofdisruption, but it will be shown later that there is not a prcise time for its onset, when it is measured with high timresolution. The erosion of the electron temperature prohas been observed in different tokamaks3,5,18,19 to have anm/n51/1 structure. We will show in this paper that thm/n51/1 erosion is precipitated by an off-axis erosion staing from the O point of them/n52/1 mode at the low fieldside ~LFS!. The EQ triggers a sequence of events thatalso common to all tokamaks, namely the fast increasepoloidal flux with its associated negative loop voltage,Vloop,and increase in plasma current,I p , and the subsequent resitive decay ofI p , known as the current quench~CQ! ~see Fig.1!.

In this paper we study how the abrupt loss of enerconfinement in RTP plasmas, preceded by am/n52/1 mode,takes place. The growth of this mode was described in dein Refs. 20 and 21. Here the discussion will focus ontime interval between the last moments of them52 growthand the time at which the plasma current starts to dec

2 © 2002 American Institute of Physics

IP license or copyright, see http://ojps.aip.org/pop/popcr.jsp

f

e

f-

3403Phys. Plasmas, Vol. 9, No. 8, August 2002 High resolution temperature and density profiles

FIG. 1. Illustration, for the cases of FPD and SPD, o~a! plasma current~solid line! and average line inte-

grated density~dotted line!. ~b! Bu at the equatorialplane at the LFS.~c! Loop voltage~solid line!, horizon-tal ~dotted line! and vertical~dashed line! plasma dis-placement.~d! Te from ECE radiometer. The circlesindicate the time of the TS profiles on the respectivFigs. 2 and 4.~e! SXR signals. The evolution ofTe(r )between the time interval limited by the two verticalgrey lines is shown in Figs. 2 and 4 for the cases oFPD and SPD, respectively. All signals are synchronized in time.

ndrie

ka

icti

,iord

b

aithr t

illons

in-lowltingtentdis-nch

ta-en

andus-

e ins iseas

ofpo-

lec-the

indicated in Fig. 1 by two vertical gray lines~Secs. III andIV !. In Sec. V we compare the EQ in minor disruptions amajor disruptions. Next, a short description of the expemental setup and of the plasmas here studied will be giv

II. EXPERIMENTAL SET-UP

The innocuous effects of disruptions in small tokamamake these very suitable for their study. Besides its smsize, RTP ~Rijnhuizen tokamak project, minor radiusa50.16m, major radiusR050.74m, Bf,2.5T. and I p

,150kA) was equipped with a set of very good diagnostthat allowed us to probe the energy quench with high spaand temporal resolution.

The time evolution ofTe(r ) with 1ms time resolution,along a horizontal chord throughout the equatorial planemeasured with a 20 channel electron cyclotron emiss~ECE! radiometer. Perpendicularly, along a vertical choTe(r ), ne(r ), andpe(r ) profiles with 3mm spatial resolutionwere measured by Thomson scattering~TS!. In order to scanthe energy quench with TS, this diagnostic was triggereda signal proportional to the amplitude ofBu,m52.

In RTP there are two types of precursors.20 The first typeis mainly characterized by an exponential increase ofBu,m52

for a short period of'0.5ms prior to disruption.21 In thesecond type,Bu,m52 stays quasisaturated for a period thcan go up to 100ms prior to disruption. Disruptions wthese two precursors will be analyzed in parallel, in orde


-n.

sll

sal

isn,

y

t

o

show their similarities and differences. Henceforth we wrefer to these disruptions simply as fast precursor disrupti~FPD! and slow precursor disruptions~SPD!, respectively. Inboth casesqa.4 with 95kA<I p<100kA and 2.08T<Bf

<2.16T. For those with fast precursors, the density wascreased by puffing Ne gas in a He plasma and for the sprecursors He gas was puffed in a He plasma. The resudisruptive events were found to be reproducible to the exthat TS profiles measured during disruptions of separatecharges could be put together to reconstruct a single queevolution.

Mode locking was never observed, i.e., the mode rotion never came to rest relative to the wall. FPD occur whthe m52 mode has a frequency of'10kHz. In SPD, thefrequency is lower, between 5 – 6kHz.

In Figs. 2~a! and 4~a! the evolution ofTe(r ) measuredby the ECE radiometer is shown for the cases of FPDSPD, respectively. With the purpose of simplifing the discsion, in the two pictures the time origin (t50) was set tocoincide with the last time them52, X point passed in frontof the ECE radiometer, before the observed fast collapsTe at the plasma core. The temperature of the isothermindicated in eV, and to enhance the contrast two gray arare used. The dark one is in the vicinity of the centralm51 mode boundary, while the lighter one is in the vicinitythe m52 separatrix facing the core. In these figures thesition of the top–bottom carbon limiter~at r51) is markedby the dotted line, to guide the eye. Figure 3 shows a coltion of TS profiles measured during an FPD and Fig. 5 is


r

le

e

.-ef

e

-

3404 Phys. Plasmas, Vol. 9, No. 8, August 2002 Salzedas et al.

FIG. 2. Illustration of a fast precursomajor disruption.~a! Time evolution,with 1ms time resolution, of the radiatemperature profile measured by thECE radiometer for the dischargshown in Fig. 1~1!, in the time win-dow between the vertical grey linesThe position of the channels is indicated at the right. The arrows indicatthe time position of the TS profiles oFig. 3. TS profile 6 is att'700ms ~seethe Appendix!. The horizontal dashedlines indicate the position of theq52surface. The dotted line indicates th

position of the limiter.~b! Bu at theequatorial plane at the LFS and displaced a toroidal angle of 150° fromthe ECE radiometer.

slereon

ts.heher

same for a SPD. The time sequence of the TS profileindicated by arrows, in Figs. 2 and 4. As said, the TS profibelong to different discharges~except for those cases that aindicated!. Their position in respect to the quench evoluti


isswas chosen with the help of ECE isotherms contour ploFor justification of the choice, the ECE contour plots of tindividual discharges are shown in the Appendix togetwith the real time points of the TS measurements.

sf

1e.-

n

FIG. 3. High spatial resolution(3mm), Thomson scattering profileobtained during the energy quench ofast precursor disruptions. Profilesand 3, belong to the same dischargThe relative time positions are schematically indicated in Figs. 1~d! and2~a!. The real time positions are showin the Appendix.


le

e

.---

e

-


FIG. 4. Illustration of a slow precursormajor disruption.~a! Time evolution,with 2ms time resolution, of the radiatemperature profile measured by thECE radiometer for the dischargshown in Fig. 1~2!, in the time win-dow between the vertical grey linesThe position of the channels is indicated at the right. The arrows 1–5 indicate the time position of the TS profiles of Fig. 5. The horizontal dashedlines indicate the position of theq52surface. The dotted line indicates th

position of the limiter.~b! Bu at theequatorial plane at the LFS and displaced a toroidal angle of 150° fromthe ECE radiometer.

le

ofhat

III. THE FIRST PHASE OF THE ENERGY QUENCHUNTIL THE FLATTENING OF Te„r …

Erosion of the central temperature profile is first visibaround t52250m s at the cold point of the centralm51


mode at the high field side~HFS! @see Fig. 2~a!#. The heatlost in the core flows to the edge leading to the expansionthe temperature profile first visible at the LFS and somewlater at the HFS. The samem/n51/1 like erosion has been

d-,e-n

FIG. 5. High spatial resolution (3mm)Thomson scattering profiles obtaineduring the energy quench of slow precursor disruptions. Profiles 2 and 3belong to the same discharge, just likprofiles 4 and 5. The relative time positions are schematically indicated iFigs. 1~2d! and 4~a!. The real time po-sitions are shown in the Appendix.


an

sie-

insm

na

ysth

e

em

n

e

ibrg

core

-

paodD

slt

he

ici--

roi

e

iei

suldrba-fast

,

tntp-tedt

nf

Forsig-e-the,

sted.of

onig.atf,

at

an-

eat

ob-

em-dius

are


observed in several tokamaks and leads to a crescent shremnant of the original hot core with circular cross sectioThe crescent becomes thinner and thinner during the eroprocess. Att5250ms, r50.5 another erosion process bcomes visible which starts to affect them52 mode by agrowing indentation of the core facing separatrix. Thisdentation describes a slight, excursion towards the placore at the angular position of them52, O point~see arrowA! and it is symmetric in time around the O point. Att50ms in the plasma core, the degradation of energy confiment intensifies, as can be seen by the widening of the dgray region atr'20.25. At the same time, the light graisothermal area expands to inside and outside and crossem52 X point, advancing towards the plasma edge atLFS.

From 0 – 40ms, as them52 mode passes in front of thECE radiometer, it is observed that erosion of theTe profileis already significant~see arrow B!. The temperature at thcold point of them51 mode decreased 50% and has becoequal to the temperature at them52, O point, at the LFS. Asseen in the first TS profile in Fig. 3, (Te1

TS) the temperatureprofile inside the hotm51 island as well as inside them52 island is irregular. InTe2

TS the profile is flattened betweethem52, X point and them51, hot point, only for positivevalues ofz/a. After this, within 40ms the temperature of thm51 hot point decreased 50% and theTe profile becomescompletely flat, up tor50.9 ~seeTe3

TS).From these observations it is seen that it is not poss

to ascribe a precise starting time for the onset of the enequench. The degradation of the energy confinement is atinuous process that becomes more intense during the pebetween250 and190ms, as shown by the excursion of thisotherms at the edge. This phase of faster erosion ofTe(r ) isobserved to start from the LFS in between them52, O pointand them51 cold point ~arrow A! proceeding asymmetrically into the plasma core, first through them51 cold pointand then to them51 hot point~arrow B!.

Despite the different evolution ofBu , a similar sequenceof events occurs for the case of the SPD, Fig. 4. One apent difference between SPD and FPD is due to the mrotation frequency that is almost two times smaller for SPIn this way, since the erosion of them51 cold point occursin approximately the same time as in the FPD ca('90ms), the fact that the mode rotates more slowly resuin an erosion that occurs within half a toroidal rotation of tm52 mode. So the beginning of them52 O point erosion~arrow A! is not always observed, since it depends on whpart of them52 mode is passing in front of the ECE radometer in this period~see, for example, Fig. 11 in the Appendix!.

From Thomson scattering, the behavior of the electdensity profile could be followed. For the FPD cases itobserved in Fig. 3 that fromne1 to ne3 the central densitydecreases'20% in the plasma core and increases'50%inside them52 island. Observing in sequence the first thrne profiles, in the region betweenz/a'0.5 toz/a'0.8, it isapparent that there is a perturbation in the density gradmovingoutwards. This indicates that the density increase


ped.on

-a

e-rk

thee

e

leyn-iod

r-e.

es

h

ns

e

ntn

the m52 island could be the result of the flow of particlefrom the island neighborhood in the core side. We shonote that such an outward displacement of density pertutions is already apparent in the last stages of theprecursor21 as is the density increase inside them52 islands.

In DIII-D ne was observed to have a similar behavior10

where an increase inne at r /a'0.75 is observed in a TSne

profile taken during the major EQ~see Fig. 2 of Ref. 10 a2.0444s). Them number, as well as the radius of the resonasurface of then51 locked mode that precedes the disrution, is not mentioned so we cannot conclude that it is relato the density increase inside them52 island observed aRTP. However, in DIII-D it was also observed that anne

decrease in the core is followed by anne increase in theplasma edge.

In the center of the profilesne2 and ne3 a step in thedensity is visible. Forne2 the highest density is betweez/a'20.5 to z/a'0, which corresponds to the hot part othem51 mode. The cause of this step is not understood.z/a.0.8 it is observed that the density does not changenificantly. So, during this period of the energy quench dspite this internal rearrangement of the electron density,global convex shape ofne(r ) is still more or less preservedas was previously observed.22

The soft x-ray~SXR! diagnostic was not tuned for thitype of discharge and often several channels were saturaIn the case of the discharge shown in Fig. 4, the intensitythe SXR radiation a few milliseconds before the disruptiwas low enough to allow the measurements shown in F1~2e!. The viewing chords of the two channels interceptr /a50.65, z/a50.75, that is, approximately the radius othe q52 surface in the LFS. The first spike in the SXRvisible on both channels, occurs when them52, X point isat the LFS. Its origin is probably related with the increasethe m52, X point temperature that occurs att50 as de-scribed before. The second spike is only visible at one chnel which means that the enhanced emissivity at them52 Xpoint dies out very quickly and asymmetrically. The sambehavior of SXR is observed for minor disruptions, and thwill be discussed in Sec. V.

IV. THE SECOND PHASE OF THE ENERGY QUENCHAFTER THE FLATTENING OF Te„r …

Shortly after the flattening ofTe(r ), i.e., when them51 erosion is completed, a series of new phenomena isserved.

The remaining gradients inTe(r ) and ne(r ) at theplasma edge disappear.

The appearance of a short-lived intense peak in the tperature and to a lesser extent in the density at a rabetween 0.7,r,0.8.

An increasing deviation betweenTe measured by TS andECE can be noticed while normally these two diagnosticsvery well in agreement with each other.


---

efy

yr.


FIG. 6. Comparison between theTeTS

profiles of Fig. 3 and theTeECE profiles

measured at the same time. It is recalled that the TS profiles are measured along a vertical chord, perpendicular to the measuring chord of thECE radiometer. Until the flattening oTe the two diagnostics agree verwell. After the flattening ofTe theelectron temperature measured bThomson scattering is always largeFor the profiles 5 the difference in thecentral temperature is already 50%.

y

r

eak-

s

A. Disappearance of gradients

The disappearance of theTe gradients at the edge is onlvisible in Te

ECE. In Fig. 2~a! ~FPD! it occurs betweent5170ms andt51100ms. The TS profile 4 does not cove


the very bottom edge of the plasma and at the top the pmasks the effect. In Fig. 4~a! ~SPD! the edge gradients disappear between120 and180ms.

After the flattening ofTe , the electron density remain

n-

.

a

FIG. 7. ~1a!–~1b! Example of a fastprecursor disruption where the erosioof the core temperature lasts for a period more than two times longer thanthe normal cases exemplified in Figs2 and 4.~2a!–~2b! Illustration of a mi-nor disruption that occurred beforeSPD.


eksito

ein

tltaS

veyT

smth

erb

m

thecoc

hen

ist,o-rum

ngurenheEfwacbp

en

, as

hattheenow-ousup-ajor

ob-on

-

theitionr

rms

eto

foret

t-

p-

eon

m-

ichea

o-owscttheeon-


high ~see Figs. 3 and 5!. However, thene profile is no longerconvex but flat or even slightly hollow. The gradients btweenr50.5 and 1 have disappeared. Except for the peane4 in Fig. 3, the other profiles show large perturbationTheir occurrence is erratic with no apparent correlation wother observations. This fact together with the large sizethese perturbations suggests that large scale turbulencducing large scale convection in the plasma may be takplace during this last phase of the energy quench.

B. The peak in Te

Atapproximately100ms an increase inTeECE at r.0.77 is

observed in Fig. 2~a!. This peak inTeECE is localized at the

same radius as them52 island. As shown in Fig. 6~4! thepeak is observed simultaneously at the top by TS and atLFS by the ECE radiometer. It was also observed simuneously at the LFS by the ECE and at the bottom by THowever, as is shown in the Appendix, this peak was neobserved by ECE at the HFS. This means that poloidallspans an angle that is between 90° and 180° at the LFS.peak value of the temperature measured inTe4

TS is two timeshigher than the average electron temperature in the placenter at the same point in time and it is 20% higher thancentral temperature before the disruption@see Fig. 9~2! in theAppendix#. The ECE radiometer measures a peak tempture of only 600eV when TS measures 1200eV. It shouldstressed that this peak is observed in all disruptions,shown in the Appendix, so it must be an intrinsic phenoenon related to disruptions.

The electron density, that was already increasing inprevious profilene3

TS in the m52 island, also peaks in thsame position as the temperature, although less pronounAnother characteristic of the peak is found in the Thomsscattering spectrum which at that place shows a pronounhigh energy tail that is not visible at other positions of tviewing chord. This indicates the presence of fast electro

Relative to the magnetic poloidal field, them52 oscil-lation is destroyed at the time the peak is observed. Thisstrong indication that them52 island is also destroyed. Buit is not clear if it is the destruction of the island that prvokes the peak or if it is the peak that provokes the desttion of the island. It should be noted that the poloidal asymetry of the peak indicates that it has nom/n51/1 topology.

C. Difference between TS and ECE

Under all normal circumstances in RTP, even durielectron cyclotron resonance heating, the ECE temperatafter correction for optical depth were in excellent agreemwith values derived from TS. As shown in Fig. 6, after tflattening of Te(r ) a major discrepancy between the ECradiometer and the TS develops, becoming as large as ator of 2, with the ECE radiometer measuring always the loest values. The measured values of the electron densitymuch lower than the cutoff value for the ECE. So this effecannot cause the observed discrepancy. The cause maydeviations from Maxwellian distribution, since the discreancy starts only after the flattening ofTe , during the onset ofthe large variations in the loop voltage.


-in.hfin-g

he-.r

ithe

ae

a-e

as-

e

ed.ned

s.

a

c--

est

ac--re

te in-

The negative spike in the loop voltage occurs betwe50 and 150ms in both examples given here@see Figs. 1~c!and 2~c!#.

In the same period the oscillations ofdB/dt are inter-rupted with a large spike, as shown in Figs. 2~b! and 4~b! andthe outer isothermal surfaces show the largest expansioncan be seen at'100ms.

V. MINOR DISRUPTION AND SLOWER MAJORDISRUPTION

The continuous degradation of energy confinement toccurs during the energy quench, up to the flattening ofTe profile, as described in Sec. III, can be interrupted. Whthis happens the energy confinement improves again alling the electron temperature to raise, returning to previvalues. An example of this process, called a minor disrtion, is shown in Fig. 7~2!. The mechanism that duringminor disruption prevents the plasma to evolve to a madisruption is not known. In this section we address this prlem heuristically, underlining the behavior of the electrtemperature in the region of theq52 surface.

To better illustrate our point, we show in Fig. 7~1! amajor disruption where the coreTe erosion lasted for a period more than two times longer than the two~typical! ex-amples shown before. In this example it is observed thatequivalent process of erosion described before at the posof arrow A, Fig. 2~a!, is preceded by a seemingly largeoutward heat flow through them52 X point, than previousexamples, as is indicated by the excursion of hot isothe@200eV and 300eV in this example in Fig. 7~1!# through them52, X point, from the core facing side30 of the island tothe edge facing side, at the LFS.

This causes them52 O point erosion, occurring at thposition of arrow A, to start at higher temperature andevolve more slowly to the plasma core. Them/n51/1 ero-sion proceeds in a similar way as the cases described be@note arrow B in Fig. 7~1!#, but the thermal energy is keplonger in the region inside theq52 radius~the light grayisothermal region only reaches the core att5220ms).

In the case of the minor disruption, Fig. 7~2!, the m52O point erosion~arrow A! is preceded also by a larger ouward heat flow through them52, X point, like in Fig. 7~1!.Such heat flow is not noticeable in the normal major disrutions @compare Fig. 2~a! and Fig. 4~a!#. Betweent50 ~arrowB! and t5150ms them/n51/1 erosion increases, but at thm52 mode it is seen that the light gray isothermal regidoes not extend further inwards~arrow C! or outwards. Af-terwards this region shrinks again and in the core the teperature at them51 mode increases.

In disruptions, them52 mode behaves like a dynamheat switch, allowing excursions of the isotherms to tedge, via the X point, and into the core, via the O point. Inminor disruption, contrary to the major disruption, the istherms excursion to the edge is interrupted and heat flmainly from the X point into the O point. This has the effeof shrinking the island and at the same time decreasingeffect that them52, O point erosion has on the core of thplasma. As the island shrinks, degradation of the energy c


s-thw-t

tin

e

nareld

ioh

ceine

de

thbeTonisditod

hes.eacn

reth

tha

ghet

cers

tionthe

y inre-

ense

thisthees

se at

out-

3Dea-

of

y

earily

xOne

ofcanple,

ec-inned

s-

-ial

tryrstes-hendes-he

tem-of

ns.eringing

ures


finement is stopped and reversed, allowing theTe profile torecover by Ohmic heating. So, from these experimentappears that them52 mode plays the critical role in increasing or reducing the rate at which energy confinement incore is degraded, as changes in the rest of the plasma aloccur after the changes in them52 island. Besides an apparent temperature dependence, it is not possible withavailable data to investigate further the structure of thisheatswitch. Nonetheless three different behaviors can be disguished.

The switchopens and closes~minor disruption!.The switchopens completely~major disruption!.An intermediate behavior between the previous on

where theheat switchdoes not open completely~major dis-ruption with longer quench time!.

VI. DISCUSSION AND CONCLUSIONS

The global scaling of the quench duration with the Ludquist number23 indicates that, in the process of nonlineinteraction between modes, the balance dominates betwthe driving force, being the tension in the magnetic fielines, and the braking force, being magnetic reconnectHowever, this balance does not explain the detailed mecnism and cannot predict in fine detail the various differenin quench behavior. A widely spread model for the beginnof the energy quench evokes the stochastization of magnfield lines between them52 and them51 modes. The non-linear interaction of these two modes would lead to thestabilization of other modes in between, mainly them/n53/2 mode, and when the width of the islands equalsdistance between them the magnetic field lines wouldcome stochastic destroying the magnetic flux surfaces.strong point for this model is that it predicts high electrheat diffusivity throughout the stochastic region, whichcomparable to the heat fluxes observed during a majorruption. Experimentally, it is not possible to measure schasticity directly and its presence has to be inferred inrectly. One expected effect is the flattening of ttemperature in the regions that have stochastic field linelimitation of this model is the lack of a description for thtime evolution of the process of stochastization and the lof a self-consistent computation of the magnetic field acurrent.

In the measurements previously described, the cleaobservation of a flattened temperature profile betweenm52 and them51 mode is inTe2

TS, Fig. 3, where it isobserved that between them52, X point at the top and them51 hot point at the bottom, theTe profile is flat. A closerlook into the same profile shows that, although at the topTe profile is flat up toz/a50.8, at the bottom there isgradient fromz/a520.45 toz/a520.7, which is the posi-tion of theq52 resonant surface. This means that the neiborhood of them52 island core facing side is not all at thsame temperature. This seems to be at variance withcomplete stochastization of this region at this advanphase of the profile erosion, as expected by some autho12

The corresponding pressure profile,pe2, is more sym-metric. It shows a small gradient atz/a50.45, which is due


it

eays

he

-

s,

-

en

n.a-sgtic

-

e-

he

s--i-

A

kd

ste

e

-

hed.

to a corresponding gradient in the density, and at the posiof theq52 surface the pressure is equal to 2kPa, both ontop and the bottom. This means that the helical symmetrthe pressure distribution inside the island is much better pserved than in temperature and density, which makes sfrom the parallel momentum balance point of view.

There are features in the measurements shown inpaper that can be interpreted as indirect indications ofexistence of electron flows. One of these indications comfrom the evolution of the density profile during theTe ero-sion, which shows a decrease in the core and an increathem52 O point, with the density perturbationmovingout-wards, opposite to the density gradient. The sequentialward displacement of thene perturbations intuitively indi-cates the existence of a flow. However, the cause andstructure of such a particle flow is not clear from these msurements. The other indirect indication of the existenceflows comes from the evolution of theTe profiles. In Figs.2~a! and 4~a!, the deformation of the isotherms indicated bthe arrow A seems to indicate the presence of an inward~i.e.,towards the core! convective motion at the O point of thm52 mode. Such a motion does not have to be necessconnected with the others described before. Linear24 andnonlinear25 MHD theory of tearing modes predicts vorteflows that advect fluid into the X point and out from thepoint. Although intrinsic to the tearing mode, these are ofluid flows. It has to be investigated if in the conditionstokamak disruptions, analogous vortex flows of electronshave the behavior described above. We note, for examthat advances in tearing mode theory26,27 predict the decou-pling between the electron and ion fluids during reconntion, which allow the electrons to play the important rolethe fast transport of energy. Such theory recently obtaisupport from experiment.28 Recent simulations of highb dis-ruptions have also found vortex flows involved in the tranport of energy during the energy quench.29

A very short-lived peak inTe is observed at every disruption. The nature of this peak is not understood. Its rad

position, as well as its correlation withBu , relates it to them52 mode. However, the reason for the helical asymmeexhibited by the peak is not clear. While in the preceding fipart of the quench, at least the helical symmetry of the prsure inside them52 island seems to be preserved, during tpeak even this is lost and it becomes doubtful if an islatopology still exists. From the Thomson spectrum, the prence of fast electrons at the time of the peak is inferred. Tdiscrepancy between Thomson and the ECE radiometerperature profiles that is observed starting at the flatteningTe is also probably due to the presence of the fast electroPossibly the destruction of them52 mode that occurs at thsame time that the peak is observed results in the scatteof the fast electrons throughout the plasma volume, inducin this way the sudden decrease inTe as measured by theECE radiometer.

After the flattening ofTe the kinetic energy is lost at acomparable rate as before the flattening, during them51erosion. In this phase, except for the peak, mode struct


in

inbe

ntm-m


are no longer observed, while the electron density remahigh.

From all these observations we note that them/n52/1,O point Te erosion associated with the density increaseside them/n52/1 island indicates that convection mayinvolved in the onset of the energy quench.


s

-

ACKNOWLEDGMENTS

F.S. was supported by the Programa PRAXIS XXI-graBD/4531/94. This work was performed under the EuratoFOM association agreement, with financial support froNWO and Euratom.

es.

-.

t

t

ere

APPENDIX: ORIGINAL DISCHARGES OF THOMSON SCATTERING PROFILES

Fast precursor disruptions

The time positions of Thomson scattering profiles from Figs. 2 and 3 are indicated here in the original discharg~SeeFigs. 8–10.!

FIG. 8. ~a! Te(r,t) as measured fromthe ECE radiometer with 1ms timeresolution. The position of the channels is shown on the right-hand side

~b! Bu(t) as measured from the fascoil with 2ms time resolution. Arrowsindicate the time position of TS pro-files 1 and 3 from Fig. 3.

FIG. 9. The same as in Fig. 8, excepthat the ECE time resolution is 10msin Fig. 1~a!. Also in this experimentsome channels of the ECE radiometwere not available, as shown on thright-hand side of Fig. 1~a!. Arrows in-dicate the time position of TS profiles2 and 4 from Fig. 3.


es.

nitial

le


FIG. 10. The same as in Fig. 8. Arrows indicate the time position of TS profiles 5 and 6 from Fig. 3.

Slow precursor disruptions

The time positions of Thomson scattering profiles from Figs. 4 and 5 are indicated here in the original discharg~SeeFigs. 11 and 12.!

FIG. 11. ~a! Te(r,t) as measured from the ECE radiometer with 2ms time resolution. The position of the channels is shown on the right-hand side. The iphase of the evolution of theTe erosion is not visible in this picture. Att550ms when them52 O point passes in front of the ECE radiometer, theTe profile

of the m51 cold point is already eroded.~b! Bu(t) as measured from the fast coil with 4ms time resolution. Arrow indicates the time position of TS profi1 from Fig. 5.

Downloaded 31 Jul 2002 to 193.136.136.136. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/pop/popcr.jsp


FIG. 12. The same as in Fig. 11 except that in Fig. 2~a!the time resolution is 20ms. Arrows indicate the timeposition of TS profiles 2, 3, 4, and 5 from Fig. 5.

alrli

9

h

naonnc

lead

ev

lead

ys

id

,on

1-5,/

s.

i,etylier,

Fu-.

at ise is

1E. P. Gorbunov and K. A. Razumova, At. Energ.15, 1105~1963!.2F. C. Schu¨ller, Plasma Phys. Controlled Fusion37, A135 ~1995!.3J. Wesson, R. D. Gill, and M. Hugon, Nucl. Fusion29, 641 ~1989!.4Equipe TFR, Nucl. Fusion17, 1283~1977!.5G. Waidman and G. Kuang,Proceedings of the 18th European PhysicSociety Conference on Controlled Fusion and Plasma Physics, BeGermany, 1991, edited by P. Bachmann and D. C. Robinson~EurophysicsConference Abstracts, Europoean Physical Society, Petit-Lancy, 19!,Vol. 15C, p. 45.

6K. H. Finken, T. Denner, and G. Mank, Nucl. Fusion40, 339 ~2000!.7O. Gruber, K. Lackner, G. Pautasso, U. Seidel, and B. Strebl, Plasma PControlled Fusion35, B191 ~1993!.

8Y. Nagayama, S. Tsuji, K. Miyamoto, K. Kawahata, N. Noda, S. Tahashi, and J. Fujita,Proceedings of the 12th International ConferencePlasma Physics and Controlled Nuclear Fusion Research, Nice, Fra1988@Plasma Phys. Controlled Nucl. Fusion Res.~IAEA, Vienna, Austria,1989!#, Vol. 1, p. 387.

9A. G Kellman, J. W. Cuthbertson, T. E. Evanset al., Proceedings of the16th International Conference on Plasma Physics and Controlled NucFusion Research, Montre´al, Canada, 1996@Plasma Phys. and ControlleNucl. Fusion Res.~IAEA, Vienna, Austria, 1997!#, Vol. 1, p. 739.

10P. L. Taylor, A. G. Kellman, B. W. Rice, and D. A. Humphreys, Phys. RLett. 76, 916 ~1996!.

11A. C. Janos, E. D. Fredrickson, K. M. McGuireet al., Proceedings of the14th International Conference on Plasma Physics and Controlled NucFusion Research, Wu¨rzburg, Austria, 1992@Plasma Phys. and ControlleNucl. Fusion Res.~IAEA, Vienna, Austria, 1993!#, Vol. 1, p. 527.

12A. Bondeson, R. D. Parker, M. Hugon, and P. Smeulders, Nucl. Fusion31,1695 ~1991!.

13B. B. Kadomtsev, Plasma Phys. Controlled Fusion26, 217 ~1984!.14P. H. Diamond, R. D. Hazeltine, B. Carreras, and H. R. Hicks, Ph

Fluids 27, 1449~1984!.15M. F. Turner and J. A. Wesson, Nucl. Fusion22, 1069~1982!.16B. V. Waddell, B. Carreras, H. R. Hicks, and J. A. Holmes, Phys. Flu

22, 896 ~1979!.


n,

1

ys.

-

e,

r

.

r

.

s

17R. Kleva and J. Drake, Phys. Fluids B3, 372 ~1991!.18J. R. Rommers, F. C. Schu¨ller, E. S. Lyadina, D. F. Da Cruz, M. Verreck

and A. J. H. Donne´, Proceedings of the 20th European ConferenceControlled Fusion and Plasma Physics, Lisbon~European Physical Soci-ety, Petit-Lancy, 1993!, Vol. 17C, Part I, 275.

19E. D. Fredrickson, K. M. McGuire, M. G. Bellet al., Nucl. Fusion33, 141~1993!.

20F. Salzedas, Ph.D. thesis, Rijksuniversiteit Utrecht, ISBN 90-3932-572000 ~URL: http://www.library.uu.nl/digiarchief/dip/diss/1940669inhoud.htm!.

21F. Salzedas, F. C Schu¨ller, A. A. M. Oomens, and the RTP team, PhyRev. Lett.88, 075002~2002!.

22F. C. Schu¨ller, A. J. H. Donne´, S. H. Heijnen, J. R. Rommers, C. P. Tanzand RTP team, inProceedings of the 21st European Physical SociConference on Controlled Fusion and Plasma Physics, MontpelFrance, 1994, edited by E. Joffrin, P. Platz, and P. E. Stott~EurophysicsConference Abstracts, European Physical Society, Petit-Lancy, 1994!, Vol.18C, p. 230.

23J. D. Callen, B. V. Waddell, B. Carreraset al., Proceedings of the 7thInternational Conference on Plasma Physics and Controlled Nuclearsion Research, Innsbruck, Austria, 1978@Plasma Phys. Controlled NuclFusion Res.~IAEA, Vienna, Austria, 1979!#, Vol. 1, p. 415.

24H. P. Furth, J. Killeen, and M. N. Rosenbluth, Phys. Fluids6, 459 ~1963!.25P. H. Rutherford, Phys. Fluids16, 1903~1973!.26D. Biskamp, E. Schwarz, and J. F. Drake, Phys. Plasmas4, 1002~1997!.27J. F. Drake, D. Biskamp, and A. Zeiler, Geophys. Res. Lett.24, 2921

~1997!.28X. H. Deng and H. Matsumoto, Nature~London! 410, 557 ~1963!.29R. Kleva and P. Guzdar, Phys. Plasmas8, 103 ~2001!.30With the core facing side is meant that side of the island separatrix th

closer to the center of the plasma. Concomitantly the edge facing sidthe side of the separatrix that is closer to the edge of the plasma.


High resolution temperature and density proﬁles during the … · 2020. 7. 29. · (3mm), Thomson scattering proﬁles obtained during the energy quench of fast precursor disruptions.

Documents