Role of Edge Current in ELM Behaviour: Modelling of Recent Current Ramp Experiment in JET

4th Transport/ITB IPTA Meeting, St Petersburg,

8-12 April 2003 1

Role of Edge Current in ELM Behaviour: Modelling of Recent Current Ramp

Experiment in JETV.V. Parail1, G. Bateman6, M. Becoulet4, G. Corrigan1, A.Y.

Dnestrovskij2, G. Huysmans4, A. Kritz6, A. Loarte5, J. Lonnroth3, T. Onjun6, G. Saibene5 , H. Wilson1

1UKAEA Fusion Association, Culham, UK; 2 Kurchatov Institute, Moscow, Russia, 3Association Euratom-Tekes, Finland; 4Association Euratom-CEA, Cadarache, France; 5 EFDA-CSU, Garching, Germany, 6 Lehigh Univ., PA

• Peeling vs. ballooning stability;• Recent current ramp-up experiments;• Predictive modelling of JET ELMy H-mode plasmas;• Conclusions and future plans


8-12 April 2003 2

Current ramp-up experiment on JETBtor=3.4T

Ip=2MA

=0.4

=1.7


8-12 April 2003 3

Peeling vs. ballooning stability (1) We adopt the following approach in predictive

modelling :

we assume certain width of the ETB (either constant or dependent on plasma parameters);

we simulate all main plasma profiles (ne, Te,i, jII, + neutrals) up to and beyond the separatrix by using a coupling of two codes: JETTO+EDGE2D;

we can make different assumptions about ELM trigger (ideal ballooning with and without access to a second stability, peeling mode);

we use ad hoc model for ELM amplitude and duration


8-12 April 2003 4

Peeling vs. ballooning stability (2)


8-12 April 2003 5


0

0.02

0.04

0.06

0.08

0.1

0 1 2 3

Stable

Unstable

J ed

ge/J

avge


8-12 April 2003 6


Simple analytical criteria for ballooning and peeling mode stability are used in JETTO as an ELM trigger:

Ballooning mode: with

adjustable parameter to account for an access to a second stability region;

Peeling mode:

critp

200

ddp

Br2

0

crit

dlBR

Bjq22

1D41 3p

2II

M '


8-12 April 2003 7


We can use either one or both criteria in our predictive modelling;

the caveat is that ELM-event itself is independent on the ELM trigger - it appears as a short sharp increase in heat and particle diffusivities (but not in electrical resistivity) within the ETB;

it means that ELM quickly destroys the edge bootstrap current but it takes much longer to dissipate the Ohmic current, which substitutes bootstrap current;


8-12 April 2003 8


We study the relative role of ballooning and peeling mode instability as ELM trigger in our predictive modelling;

In experiment the easiest way to decouple the edge current from the pressure gradient (bootstrap current) is to ramp the current during the ELMy H-mode;

JET recently conducted a dedicated experiment, in which current ramp-up technique was used in low density Optimised Shear plasmas;


8-12 April 2003 9

Selection of shots (1)Btor=3.4T

Ip=2MA

=0.4

=1.7


8-12 April 2003 10

Selection of shots (2)


8-12 April 2003 11

Predictive Modelling of ELMs (1)First we demonstrate that we can reproduce

experimental data with ballooning stability trigger;

Ballooning stability trigger with

JET #53298

2crit


8-12 April 2003 12

Predictive Modelling of ELMs (2)First we demonstrate that we can reproduce

experimental data with ballooning stability trigger;


8-12 April 2003 13

Predictive Modelling of ELMs (3) Then we study the relative role of ballooning and

peeling mode as an ELM trigger;


8-12 April 2003 14

Predictive Modelling of ELMs (4) Or with a better resolution:


8-12 April 2003 15

Predictive Modelling of ELMs (5)

We conclude that both ballooning and peeling mode trigger ELM within the experimentally observed parameter range;

If peeling mode is used as ELM trigger the ELM duration becomes much longer than experimentally observed length telm<0.2msec;

The explanation is that the edge current should be removed in order to suppress peeling mode and this current is controlled by the skin time;


8-12 April 2003 16


We also found that ELM has a composite character in many cases: it is triggered by ballooning mode, which is superseded by the peeling mode:


8-12 April 2003 17



8-12 April 2003 18

Predictive Modelling of ELMs (8) It is interesting to note that since skin time is

proportional to skinTe3/2, the ELM duration

should depend on edge temperature;


8-12 April 2003 19

Predictive Modelling of ELMs (9) Now we can check how our combined peeling-

ballooning model for ELM trigger will react on the current ramp:


8-12 April 2003 20

Predictive Modelling of ELMs (10) or, with a better resolution:


8-12 April 2003 21

Predictive Modelling of ELMs (11) Qualitatively (only!) it goes in line with experiment:


8-12 April 2003 22

Predictive Modelling of ELMs (12) Qualitatively (only!) it goes in line with experiment;

We can easily reproduce temperature and density profiles but not ELM dynamics;


8-12 April 2003 23

Conclusions

Experimental observations and predictive modelling show that both ballooning and peeling mode can play their part in ELMs;

Recent current ramp-up experiment in JET (although not quite conclusive) indicate that peeling mode might play a crucial role in keeping plasma edge in type-III ELMy H-mode in low density plasmas;


8-12 April 2003 24

Conclusions

Role of Edge Current in ELM Behaviour: Modelling of Recent Current Ramp Experiment in JET

Documents

ballooning mode

current peeling

mode stability

mode trigger elm

elm trigger ideal ballooning

stability region

elm behaviour

elm amplitude