Evans-ELMs by xiangpeng

VIEWS: 5 PAGES: 14

									 Edge Localized Mode and pedestal control
 using resonant magnetic perturbations                                 T. E. Evans
• Transient heat flux excursions are a critical ITER issue: ITER 2007-20162037
    –   Transient energy impulses to material surface must
        be < 45-60 MJm-2s-1/2 (ideal ablation limit)
        •   Implies transients (ELMs ~0.5 ms) < 6-8 MJ (assuming in-
            out and axisymmetric target distribution)
        •   Scaling from present tokamaks~12-20 MJ
•   In DEMO transients heat flux excursions are expected to
    be at least 5X large than in ITER
•   A reliable ELM control system is essential for DEMO
                                                                       DEMO 2024-2036
•   Resonant Magnet Perturbations (RMPs):
    –   Produce complete ELM elimination
        •   at reactor relevant collisionalities
        •   with robust edge transport barriers and
        •   somewhat improved Teped
    –   Appear to be scalable to reactor plasmas
    –   Provide pedestal and steady-state heat flux control
        that may lead to improved H-mode performance

                                                                             tee-07fsc-1/2
RMP ELM control in DEMO requires developing
scalable physics models and optimized coil designs
•   Current RMP experiments are providing valuable basic physics data:
    –   Current physics understanding  large pedestal resonances with minimal core
        resonances and non-resonant components
        •   Currently restricted to suboptimal coils (using field-error correction and RWM control coils)
•   ITER attempting to shoehorn RMP ELM control coils into its design but facing:
    –   Rigid constraints on coil design parameters
        •   Suboptimal design, reduced spectral flexibility
        •   No provisions for mid-course coil optimization

•   The program is facing a significant gap in the development of RMP ELM
    control for DEMO that will not be filled by ITER
•   This gap could be filled with a dedicated new device or a significant
    upgrade (~100M$) of an existing device
    –   Optimized RMP coil design tightly integrated into device design
        •   Options for mid-course coil and divertor changes based on developing physics
            understanding ( strongly integrated theory, modeling and experiments)
    –   New physics understanding  potential for improved DEMO performance

                                                                                        tee-07fsc-2/2
Supporting material




                      tee-07fsc-3/2
ELMs are completely eliminated with RMPs in high
confinement plasmas with ITER Similar Shapes




•   2006 lower divertor reconfiguration allows collisionality control (pumping)
    in ITER Similar Shape

                     T.E. Evans, et al., Nature Physics, 2 (2006) 419.
                                                                         tee-07fsc-4/2
The predicted tangle forms non-axisymmetric magnetic
footprints which have been experimentally observed

    123300: filtered CIII Xpt-TV




    123301: filtered D Xpt-TV




•   Te reflects a superposition of both upper invariant manifolds
•   Multiple footprint stripes observed during I-coil RPM operation

                    I. Joseph, et al., Nucl. Fusion, (2007) to be submitted.
                                                                               tee-07fsc-5/2
3D structure of tangle is seen by rotating the
magnetic perturbation toroidally




                             QuickTime™ and a
                                decompressor
                       are need ed to see this picture.




             I. Joseph, et al., Nucl. Fusion, (2007) to be submitted.
                                                                        tee-07fsc-6/2
The peak divertor heat flux is reduced by 40% due to
separatrix splitting when the RMP is applied




   •   Split heat flux peaks are consistent with divertor plate
       homoclinic tangle intersections

                T.E. Evans, et al., J. Physics: Conf. Ser., 7 (2005) 174
                                                                           tee-07fsc-7/2
Installation of magnetic coils on MAST

   Plan to install 12 “DIII-D I-coil” in MAST



                                                                To be used for
                                                                ELM mitigation
                                                               and TAE studies




       Complete installation by the end of 2007

         A. Kirk, et al., 3rd Stochasticity in Fusion Plasmas (2007), Juelich, Germany.
                                                                              tee-07fsc-8/2
Proposed RMP coil design for ASDEX-U




   •    Both n=3 and n=4 RMP ELM control experiments will be possible
   W. Suttrop, et al., 3rd Stochasticity in Fusion Plasmas (2007), Juelich, Germany.
                                                                                       tee-07fsc-9/2
Several RMPs coil design options are being studied
for ELM and heat flux control on ITER




   •   Both internal (n=3 and n=4) and external (n=3) RMP ELM and divertor heat
       flux cotrol coil design are being modeled for ITER
                 M. Becoulét, et al., Nucl. Fusion (2007) submitted.
                                                                       tee-07fsc-10/2
RMP H-modes have reduced particle confinement times
compared to ELMing H-modes




•   Three small ELM-like D
    bursts are triggered during
    the HFS pellet ablation phase
    but do not persist.


                  T.E. Evans, et al., Nucl. Fusion (submitted June, 2007)
                                                                            tee-07fsc-11/2
RMP ELM suppression is correlated with a narrowing and
shifting of the pedestal gradient profile




             T.E. Evans, et al., Nucl. Fusion (submitted June, 2007)
                                                                       tee-07fsc-12/2
Calculations of the stochastic layer width are
sensitive to plasma current and pressure profiles




                 Spectral gap
               Spectral gap


     Island width




•   Fast ion pressure (measured) and edge bootstrap current distribution (modeled) are needed for
    accurate calculations of the stochastic layer width.




                                                                                 tee-07fsc-13/2
RMP induced edge transport barriers have been
observed in limiter tokamaks




                                                    Confirmed in TEXTOR: K. H. Finken,
                                                    et al., Phys. Rev. Lett. 98 (2007) 065001
            T.E. Evans, et al., J. Nucl. Mater. 196-198 (1992) 421
                                                                          tee-07fsc-14/2

								
To top