c-mod pedestal program · 2011. 4. 20. · alcator c-mod pac meeting: feb 4—6, 2009: pedestal...

C-Mod Pedestal Program

Alcator C-Mod Program Advisory Committee Meeting

February 4—6, 2009Presented by Jerry Hughes

Alcator C-Mod PAC Meeting: Feb 4—6, 2009: Pedestal Slide 2 of 26

Pedestal physics: An integral component of the C-Mod research program

• Research highlights and plans covered in this presentation:– Pedestal structure and transport– Edge relaxation mechanisms– L-H thresholds and transition physics– Pedestal control– Theory and simulation

• Edge barrier formation, profile structure and relaxation processes all play critical roles in high-performance operation of tokamaks

• Issues permeate number of topical science areas and programmatic thrusts, and research contributes to FESAC, ITER priorities

• Ultimate goal: physics-based models for burning plasma which are scalable to ITER and beyond

• C-Mod occupies a unique parameter space that complements studies on other devices (large B/R, neL, range of pedestal collisionality)


C-Mod exploits large range of operational space for pedestal studies

• C-Mod pedestal studied over extended range of engineering parameters: e.g.BT, IP, ne

• Increased studies of operation with “alternative” magnetic topology– Extremes in shaping– Unfavorable ion B drift direction (in

both normal and reversed BT direction) – Near double null

• Lower collisionality with above techniques, cryopumping

• Access a wide range of regimes– EDA, ELMy H-modes– Improved L-mode with T pedestal

Typ. C-ModITER-likeJFT-2M-like


C-Mod exploits large range of operational space for pedestal studies

• C-Mod pedestal studied over extended range of engineering parameters: e.g.BT, IP, ne

• Increased studies of operation with “alternative” magnetic topology– Extremes in shaping– Unfavorable ion B drift direction (in

both normal and reversed BT direction) – Near double null

• Lower collisionality with above techniques, cryopumping

• Access a wide range of regimes– EDA, ELMy H-modes– Improved L-mode with T pedestal


We are enabled by an extensive set of well-resolved edge diagnostics

• Thomson scattering (Te, ne)• CXRS (TI, vIθ, vIφ)

– Inner wall toroidal views (passive and gas-puff assisted)

– Pedestal beam-based CXRS (toroidaland poloidal views)

• Scanning Mach probes, HFS+LFS (Te, ne, v)

• Electron cyclotron emission (Te)• Visible bremsstrahlung (neZeff

1/2)• Soft x-rays (nI)• Neutral emissivity measurements

(passive, gas puff imaging)• Reflectometer (ne fluctuations)• Phase-contrast imaging (ne flucts.)

Pedestal diagnostic set emphasizes millimeter-resolution profiles, fluctuations

Tor.

CXRS

(GPI)

CXRS

(DNB)

Pol/tor

Verticalscanningprobe

Edge/SOL

Thomson

Horizontalscanningprobe

HFS Wallscanningprobes

Toroidally

viewing

Lyα arrays

Toroidally

viewing

Lyα arrays

Select edge diagnosticsfor flow and neutral studies

= Existing= Planned

Select edge diagnostics


We are enabled by an extensive set of well-resolved edge diagnostics

• Thomson scattering (Te, ne)• CXRS (TI, vIθ, vIφ)

– Inner wall toroidal views (passive and gas-puff assisted)

– Pedestal beam-based CXRS (toroidaland poloidal views)

• Scanning Mach probes, HFS+LFS (Te, ne, v)

• Electron cyclotron emission (Te)• Visible bremsstrahlung (neZeff

1/2)• Soft x-rays (nI)• Neutral emissivity measurements

(passive, gas puff imaging)• Reflectometer (ne fluctuations)• Phase-contrast imaging (ne flucts.)

Pedestal diagnostic set emphasizes millimeter-resolution profiles, fluctuations

Distance from Separatrix (cm)

Pol

oida

l Rot

atio

n (k

m/s

)L-ModeImproved L-ModeH-Mode

Tem

pera

ture

(eV

)

Te (Thomson)Ti (CXRS)

1070726014

Student involvement in diagnostic development is high


Pedestal structure and transport: Edge flows and radial electric field• Edge CXRS is providing excellent results

– TI~Te over a wide range in collisionality– Strong vθ shear in H-mode– Strongly negative Er well develops

following L-H transition; depth correlates with confinement

– With results from other devices, consistent with a R0 scaling for Er well width

• Research plans:• Relate transport reduction, fluctuation

suppression, confinement to level of ExBshear in the pedestal region in various confinement regimes

– Explore decoupling of particle and thermal transport suppression (improved L-modes)

• Cross-machine comparisons of Er well– Relative contributions of flow, diamagnetic

components– Revisit non-dimensional pedestal matching

experiment between C-Mod/DIII-D• Momentum transport through pedestal understanding intrinsic rotation without

external torque– What is the source of large H-mode toroidal rotation?– How are the core rotation and SOL flows coupled?

Total Radial Electric Field

Distance From LCFS (cm)

Ele

ctric

Fie

ld (

kV/m

)

L-mode

EDA H-mode, H89~1.5

ELM-free H-mode, H89~2.2

R. McDermott, PhD student


Pedestal structure and transport: Edge flows and radial electric field• Edge CXRS is providing excellent results

– TI~Te over a wide range in collisionality– Strong vθ shear in H-mode– Strongly negative Er well develops

following L-H transition; depth correlates with confinement

– With results from other devices, consistent with a R0 scaling for Er well width

• Research plans:• Relate transport reduction, fluctuation

suppression, confinement to level of ExBshear in the pedestal region in various confinement regimes

– Explore decoupling of particle and thermal transport suppression (improved L-modes)

• Cross-machine comparisons of Er well– Relative contributions of flow, diamagnetic

components– Revisit non-dimensional pedestal matching

experiment between C-Mod/DIII-D• Momentum transport through pedestal understanding intrinsic rotation without

external torque– What is the source of large H-mode toroidal rotation?– How are the core rotation and SOL flows coupled?

1.0

1.5

2.0

2.5

3.0

0 100 200 300H

89

EDA

ELM-free

Er Depth (kV/m)

R. McDermott, Phys. Plasmas(forthcoming)


Pedestal structure and transport: Pedestal width physics

• Pedestal width nearly invariant under typical C-Mod operating conditions (EDA/ELM-free)

– No apparent scalings with ρφ, ρθ, βpol

– Does not scale with neutral fueling depth– Indications that magnetic shear plays a

role – Proximity to double null equilibrium can

impact attainable width (more later)• Width scaling in Type I ELMy H-mode is

consistent with βpol1/2 scaling, consistent

with DIII-D results

• Research goals• Further explore dependence of pedestal structure (width, gradients) on magnetic

topology and shape • Assess the role of magnetic shear in setting Δ• Collaboration with DIII-D, NSTX to refine width scaling, compare with models

(proposed FY11 Joint Research Milestone)

DIII-D scaling

In collaboration with P. Snyder (GA)


Pedestal structure and transport: Gradient limits in edge transport barriers

• Processes that limit pedestal gradients between or in the absence of ELMs are important for determining pedestal structure

– On C-Mod, normalized pressure gradient αMHD (~∇p/IP2) decreases with increasing collisionality ν*

– Similar “critical-gradient”phenomena seen in near SOL in L-mode, along with substantial evidence of gradient-driven radial flux

• Research Plans:– Does pressure gradient saturate at lower collisionality? (ELM-dominated?)– Develop physical understanding of processes that set flux-gradient relationships

in the absence of (or between) ELMs important for FY11 milestone– Examine the pedestal “phase-space” in alternate configurations (e.g. DN,

reversed BT, extreme triangularity) – How does this connect to L-mode critical gradient results in Ohmic plasmas?

21 10 70

50

10

QCMno QCM

Mid-pedestal ν*

Mid

-ped

esta

l αM

HD

Expansion toward lower collisionality


Pedestal structure and transport: Fueling and H-mode density control

• Density pedestal height determined to leading order by plasma current and otherwise challenging to control

• Density gradients are stiff to changes in fueling via puffing, pumping, though collisionality can be modified

• Research goals and plans:• Ongoing studies of profile stiffness,

pedestal width, are being extended to lower collisionality

• Enhance the 2D picture of ionization source inputs to modeling

– Take advantage of neutral emissivity profiles at multiple poloidal locations

– Modeling being done to facilitate interpretation of these measurements

const. p curves ~ Ip2

ne,PED (1020m-3)T e

,PED

(keV

)

USN, normal fieldw/ Ptot = 3.8 +/- 0.2 MW

1.0MA

0.6MA

unpu

mped

pumpe

d

0.8MA Improved H-mode confinement with active pumping

• Examine H-mode fueling of pedestal at ITER BT, q as ITER neutral opacity is approached

• Opportunities for pedestal optimization and control are being explored as well (more later)


Edge relaxation mechanisms: Continuous pedestal regulation

• Objective: Understand the physical processes determining the operational space of edge relaxation mechanisms in H-mode

– Are small-ELM or no-ELM regimes compatible with a high-confinement ITER pedestal?

• EDA H-mode: Continuous transport process driven by fully electromagnetic quasi-coherent mode (QCM) in pedestal

– Usually with no ELMs– Compatible with the appearance of small ELMs

at sufficiently high power

• Research goals and plans: Understand the natural QCM drive via additional experimental diagnosis and theory/modeling

– Extensive measurements available: PCI, magnetics, reflectometry, fast Dα

– Probe with radially spaced elements into Ohmic EDA H-mode to determine radial extent and position of the QCM

– Test consistency with resistive ballooning using BOUT code (Umansky, LLNL)– Evaluate other candidate mechanisms (e.g. K-H instability, saturated kink/ballooning mode)

PCI

0.61.0

1.4

1.8

Line-averaged ne

time (s)


Edge relaxation mechanisms: Edge-Localized Modes

• Large Type I ELMs studied in atypically shaped discharges (δlower>0.75, δupper~0.15)

• Small ELM comparison with NSTX, MAST examined access conditions, structure of small ELMs

– Accessibility of small ELMs improved in near double null discharges

– Pedestal and confinement also improved • Goals and plans:• Continue studying structure and dynamics of

edge filaments in ELMs of varying size • Resolve boundaries of relaxation regimes in

operational space• Theoretical understanding of the ELM

stabilization obtained naturally on C-Mod (P. Snyder, GA)

• Focus on role of shape: How does shape affect underlying pedestal transport, ELM stability?

• More efficient stability analysis through importation of data handling codes (T. Osborne, GA)

• Use stability codes to aid in ELMy discharge development

small ELMsno ELMs

SSEP~-5mm

In collaboration with NSTX, MAST


L-H transition physics: Ongoing and future contributions to ITER

• Auxiliary power required for H-mode access a critical issue for ITER mission

– C-Mod is participating in an ITPA task force to address projection of H-mode thresholds

• Scaling of low-density limit for L-H transition is an ITER priority

– C-Mod nLH,min almost 2x higher than planned ITER target density Worry?

– ITPA joint experiment among C-Mod, DIII-D, JET, AUG, TCV

• Demonstrated that nLH,min does not scale with IP• BT dependence checked with both Ohmic and

ICRF-induced L-H transitions • Ohmic results appeared to show a ne limit

strongly decreasing with field• BT sensitivity not so great in ICRF heated cases

• ITER requires a He phase for system commissioning

– Will assess H-modes in He plasmas (L-H transition power, confinement properties, comparison of edge relaxation)


L-H transition physics: Plans

• Understand role of SOL flows, edge rotation shear in ETB formation, through more routine diagnosis of pedestal rotation velocities, Er

• Continue study of unfavorable Bdrift discharges • 2-phase transitions to H-mode,

showing gradual suppression of fluctuations, Te evolution, followed by L-H bifurcation

• Slow evolution of edge profiles, flows favorable for resolving L-H transition trigger

• If L-H transition is suppressed, can lead to an improved L-mode

Upper NullDouble NullLower Null


Improved L-modes: an alternative to traditional H-mode

• Suppression of traditional H-mode: Good for burning plasma operation?

• Obtained in high power discharges with unfavorable B drift (high PLH)

• Low particle confinement combined with high energy confinement (H98~1)

• Temperature pedestals of ~1keV have been obtained with benign impurity confinement, steady

• No ELMs• Observation of broad EM mode in edge

– Similar characteristics to QCM associated with EDA

– Likely relaxation mechanism for density gradient

– But, favored by low ν*

3.0

Time (s)0.6 0.8 1.0 1.2 1.4 1.6

150250

50100200

0.80.4

2

43

1.02.0

0.51.5

21

3

0

2.0

PRAD (MW)

Density (1020m-3)

H98

RF Power (MW)

I-mode

1.0

Magnetic Fluctuations

0.60.2

2.5

1.0

D-α

Temperature (keV)

2

43

1

1

Elm-free H-modes


Improved L-modes: Plans

• The pedestal in an improved L-mode is at a pressure pedestal and collisionality similar to that of ELMy H-mode

– Comparative stability analysis– Examine conditions for existence

of continuous modes is there a connection to edge harmonic oscillations on DIII-D?

• Optimization of shape, other parameters to keep H-mode suppressed at higher power

• Improved L-modes are attractive targets for research requiring low density, good confinement

– Core density peaking– LHCD experiments– Advanced scenario research

Steady EDA

discrete ELMs

1000

800

600

400

200

0T e

ped (e

V)

neped (1020 m-3)

1 20 3


Improved L-modes: Plans

• The pedestal in an improved L-mode is at a pressure pedestal and collisionality similar to that of ELMy H-mode

– Comparative stability analysis– Examine conditions for existence

of continuous modes is there a connection to edge harmonic oscillations on DIII-D?

• Optimization of shape, other parameters to keep H-mode suppressed at higher power

• Improved L-modes are attractive targets for research requiring low density, good confinement

– Core density peaking– LHCD experiments– Advanced scenario research

1

1.5

2

2.5

3

3.5

4

1 1.5 2 2.5 3 3.5 4

Ip = 0.8 MA, q ~ 4.71:1Ip=1.35 MA, q ~ 2.9

Pth,scaling = 0.628 x 1.1 δκ 0.28 ne,av1.1

Mea

sure

d P

th (M

W)

Nominal PLH with favorable grad-B drift


Pedestal optimization and control is a developing priority for C-Mod

• H-mode density control challenging due to stiff pedestal transport

• Developing shapes favorable to low-ν*, high confinement H-modes, in near steady state

– Improved L-mode – Near double null EDA H-mode

• Pedestal shows strong sensitivity to magnetic balance when the distance between magnetic separatrices (SSEP) becomes comparable to pedestal width

– Important for ITER?• Will use magnetic balance as a

knob for obtaining target pedestal parameters.

– Lower density for SSEP~+5mm– Best confinement obtained for

SSEP~-5mm. Small ELMs readily seen

n e,P

ED (1

020m

-3)

1

2

3

4(a)

Typ. LSN range

Typ. USN range

SSEP (cm)0 1-2 2-1

T e,P

ED (k

eV)

0

(b)

0

.2

.4

.6

.8

1.0

PNET : 2.5—3.0MWPNET : 2.0—2.5MWPNET : 1.5—2.0MWPNET : 1.0—1.5MWPNET : 0.0—1.0MW

PNET = PTOT - dWP/dt -PRAD

Open diamonds:Individual time points

Closed symbols:binned and averaged


Pedestal modification with lower hybrid

• Combining LHCD with low-density H-modes gives up to 30% reduction in pedestal density

• Temperature increases with the result the pressure pedestal is the same or higher

• Transient behavior of edge, core plasma suggest transport is being altered in stages, possibly at different radial locations

– Application of LHRF results in a prompt effect on scrape-off layer, with core, pedestal effects following

• Plans:• Improve modeling of LH wave propagation• Understand damping and determine balance

of heating/CD• Compare results with the application of

ECH/ECCD on other devices• Explore LHRF as a tool for controlling

pedestal structure, ELM quality

T e (k

eV)

p e (k

Pa)

R - RLCFS (mm)

n e (1

020 m

-3)

Edge pedestal: LH off / LH on


Pedestal modification with lower hybrid

PCI Channel 20Shot 1080306013

@LCFS

SOL

Inside pedestal

R=65.8cm

Approx. beginning of grad-n relaxation

• Combining LHCD with low-density H-modes gives up to 30% reduction in pedestal density

• Temperature increases with the result the pressure pedestal is the same or higher

• Transient behavior of edge, core plasma suggest transport is being altered in stages, possibly at different radial locations

– Application of LHRF results in a prompt effect on scrape-off layer, with core, pedestal effects following

• Plans:• Improve modeling of LH wave propagation• Understand damping and determine balance

of heating/CD• Compare results with the application of

ECH/ECCD on other devices• Explore LHRF as a tool for controlling

pedestal structure, ELM quality


Theory and simulation: Plans

• Use measurements to test theoretical predictions of edge Er and its role in pedestal formation

• Use computational tools to enhance our physical understanding, and contribute data for validation of newest edge codes

• Simulate edge turbulence using BOUT (M. Umansky, LLNL)• Continue ELM/EDA studies with ELITE, M3D• Utilize XGC0 for pedestal transport calculations, with 3D EM

turbulence calculated by XGC1 (C.-S. Chang, NYU) • Work closely with Center for Plasma Edge Simulation

– Code validation– Integrated work flow for simulating complex time-dependent edge

phenomena (ELM cycle, L-H transition)• With DIII-D, NSTX, we will work to test models for pedestal structure

as part of a proposed 2011 FES Joint Research Milestone


We are positioned to contribute to a number of ITER priority tasks

• Improve predictive capability of pedestal structure– Cross machine comparisons to isolate physics setting pedestal width– Utilize profile database for integrated modeling of pedestal structure and

transport comparison to experiment– Establish pedestal profile database for hybrid and advanced regimes– Assess impact of ELM control techniques on pedestal structure

• Improve predictive and design capability for small ELM and quiescent H-mode regimes and ELM control techniques– Assess applicability of low collisionality small ELM regimes– Test nonlinear MHD and turbulence models of ELM evolution

• Re-examine L-H power threshold at low density• Assess H-mode access, pedestal and confinement properties in He

plasmas• Examine core fueling efficiency at neutral opacity approaching that

of ITER


We participate in ongoing inter-machine collaborations through ITPA

Description ITPA designation

Notes on C-Mod contributions

Pedestal structure and ELM stability in double null

PEP-6 H-modes in near DN and SN configurations are compared in terms of profile structure and ELM stability.

Small ELM regime comparison on C-Mod, NSTX and MAST

PEP-16 High-power ELMy regimes accessed in double and single null configurations

Controllability of pedestal and ELM characteristics by edge ECH/ECCD/LHCD

PEP-22 Modifications to pedestal observed with application of LHRF. Will examine effects of edge CD, electron heating on pedestal transport, ELM stability

Scaling of the Low-Density Limit of the H-mode Threshold

TC-3 Provided high-field data on current and field scaling of low-density limit. Evaluating impact of wall conditions, character of edge fueling


We are likely to participate in additional joint experiments

Description ITPA designation

Basic mechanisms of edge transport with resonant magnetic perturbations in toroidal plasma confinement devices

PEP-19

Documentation of the edge pedestal in advanced scenarios PEP-20

Power ratio – Hysteresis and access to H-mode with H~1 TC-2

H-mode transition and confinement dependence on ionic species

TC-4


In conclusion

• C-Mod makes valuable contributions to pedestal physics relevant to burning plasmas and ITER development– L-H transition physics– Barrier structure (width, gradient scalings)– Edge relaxation mechanisms

• Pedestal program priorities– Improving experimental diagnosis of pedestal profiles, fluctuations,

edge flows– Pedestal studies in an extended range of machine parameters,

equilibrium configurations – Seeking better understanding of transport, edge stability through

modeling, simulation – Collaboration with other facilities to develop multi-machine scalings– Optimization and control of pedestal in various confinement regimes– Support of integrated scenario development

c-mod pedestal program · 2011. 4. 20. · alcator c-mod pac meeting: feb 4—6, 2009: pedestal...

Documents