1 of 27

Exploring Negative Triangularity Through Modeling and Experiment

A.O. Nelson, C. Paz-Soldan, H. Wilson, S. Guizzo, S. Freiberger, �L. Rondini, M. Pharr, N. Logan, C. Hansen

with

J. Parisi, X. Zhang, H. Anand, W. Wehner �and the DIII-D NT Team

APPLIED PHYSICS AND

APPLIED MATHEMATICS

WITH MATERIALS SCIENCE & ENGINEERING

COLUMBIAENGINEERING

The Fu Foundation School of Engineering and Applied Science

1

2 of 27

This talk will run through Columbia activities under DE-SC002272

  1. Brief highlights from the DIII-D NT campaign
  2. Assessment of H-mode inhibition for NT plasmas
  3. Computational exploration
  4. Validation on DIII-D dataset
  5. Ongoing efforts to realize NT on MAST-U
  6. Initial performance scoping
  7. Development of vertical control algorithms
  8. Scoping of NT reactor performance
  9. MANTA
  10. Highly radiative NT scenarios
  11. Bonus: Highlights from Columbia efforts on resistive stability

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

2

3 of 27

DIII-D NT campaign achieved high performance without ELMs

  • DIII-D completed dedicated NT campaign in Jan/Feb 2023
  • Tiles installed on the outer wall allowed for high power, diverted plasmas
  • Highlights include:
  • Robust ELM-free operation
  • Simultaneous access to � H98>1, βN>2.5 and q95<3
  • Dissipative edge solution obtained with� divertor detachment
  • Reliable MHD and vertical stability
  • Greenwald fractions of fGW≲2

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

3

4 of 27

Infinite-n ballooning mode gives upper bound for NT edge

  • BALOO is used to model infinite-n ballooning stability in NT edge
  • Access to 2nd stability closed at δ < δcrit
  • δcrit depends primarily on geometry through local magnetic shear

  • This sets an upper bound on the NT pedestal below* the PB limit

exp. 𝛂

ballooning limit

H-mode!

L-mode!

[A.O. Nelson et al 2022 Nucl. Fusion 62 096020]

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

4

5 of 27

DIII-D NT plasma consistent with ideal ballooning limit

  • During DIII-D campaign, no ELM were encountered past δ ≲ -1.5
  • Agrees exactly with closure of the 2nd stability limit

  • High power, high performance, �detachment all compatible �with ELM-free operation

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

5

6 of 27

ELM-free NT scenarios should also be accessible in STs!

  • Aspect ratio changes the critical δ at which 2nd stability access is closed
  • In STs, δcrit is lower
  • In STs, ballooning limit for pped is lower

  • On NSTX-U and MAST-U, preliminary modeling shows that the access to �2nd stability region can be closed!

  • STs can typically also access H-modes�in the 1st stability region…
  • Does this still apply in NT?

[Nelson et al 2022 NF 62 096020]

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

6

7 of 27

Scoping NT plasmas on MAST-U: coli constraints satisfied

  • Conservative options available with reduced current limits �
  • On right:
    • Ip = 0.6 MA
    • δ = -0.1
    • no coil at more than 70% of real limit�
  • Further optimization for initial test possible

coil limit

coil request

7

8 of 27

Full orbit calculations show enhanced fgc in MAST-U NT

  • Orbit tracking simulations show
  • small enhancement in ftrapped with NT
  • large enhancement in fg.c. with NT

  • Effect strongest at edge:�almost 40% g.c. trapped!�
  • Has a less destabilizing effect on turbulence… � …linked to confinement� improvement!

Work done by students L. Rondini and S. Freiberger

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

8

9 of 27

STEP scan shows increasing performance in NT

rotation fixed

decreasing triangularity

fixed heating power

Work done by graduate student �H. Wilson

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

9

10 of 27

Vertical stability still an open question for NT reactors

  • NT plasmas typically subject to larger vertical instability growth rates

  • Evident on DIII-D campaign with unoptimized coil sets
  • Dataset used to validate TokSys

  • Using new software (TokaMaker) to explore options for NT reactors
  • Non-conformal wall significantly improves vertical stability
  • Coil sets can be �optimized to �increase control

[A O Nelson et al 2023 PPCF 65 044002]

Work done by undergraduate student S. Guizzo

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

10

11 of 27

Development of control schemes on MAST-U is ongoing

  • Initial experiments scheduled for mid-September!
  • Start with low current, Ohmic plasmas
  • World first SNT!

  • Goals:
  • Achieve NT on MAST-U
  • Test NT vertical stability on STs
  • Measure basic confinement dependence

  • Working closely with GA team to develop NT shape controllers for MAST-U
  • Control simulations based in TokSys
  • Significant help from H Anand, W Wehner

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

11

12 of 27

CU+MIT graduate students teamed up to design MANTA

  • Large aspect ratio, low elongation, and NT reduces physics and technology risks

  • Variable fusion power at constant, manageable divertor power exhaust

  • Pulsed, but constant electricity production > 50 MWe & Qelectrical>1

  • Tritium breeding beyond �sufficient for operation and �inventory (TBR ≳ 0.9)

  • Overnight cost < $5B

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

12

13 of 27

STEP enables for self-consistent core modeling + radiators

  • STEP allows for integrated modeling of arbitrary devices by connecting utility codes:
  • CHEF: heating and current drive
  • CHEASE: equilibria
  • TGYRO: transport and profiles (variable SAT rules)

  • Implementing edge model!
  • EPED for PT
  • BALOO for NT (+gkped??)

  • Exploring the use of radiators!
  • Validate modeling with high frad DIII-D
  • Extend analysis to MANTA and other�future machines

Work done by graduate student �H. Wilson

radiator: Ne

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

13

14 of 27

Looking forward: using gkped to determine edge constraints

  • What physics actually sets the NT edge profiles?
  • Use gk_ped to investigate the role of �gyrokinetic instabilities (KBM, ETGs, etc.) �in limiting NT edge T & n gradients
  • Validate with data from DIII-D, MAST-U

  • How does NT (and reduced 2nd stability limit)�interact with increased β in STs?
  • Is NT performance enhanced or hindered�by ST operation?
  • Prediction and validation for SMART

  • How does NT physics scale towards reactors?
  • Assembling international database for comparison with modeling!

[Courtesy of J. Parisi]

ballooning critical pedestal

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

14

15 of 27

High ITER coil tilting/shifting risk and CC extrapolability

14.9%

18.4%

3.0%

1.2%

Locking risk in IBS without EFC

Predicting correction current extrapol- ability from low power scenarios to IBS

Work done by graduate student �M. Pharr

Shifting + tilting PF, CS coils within tolerances and using linear poloidal mode coupling physics (GPEC) to find dominant resonant error field

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

15

16 of 27

– Backup Slides –

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

16

17 of 27

In single gyrokinetic study, KBMs limit NT ST performance

  • In NT, second stability window DNE → core is ballooning unstable
    • Strong KBMs that limit transport → reduced β limit!

  • Turbulence challenges reduced at high field

R. Davies, et. al., (2022)

PPCF, 64 105001

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

17

18 of 27

DIII-D NT armor

New Armor Tiles

New Pedestals

R-1

R-2

18

19 of 27

DIII-D ELM-free phenomenology robust across database

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

19

20 of 27

More NT pulling coils on MAST-U than DIII-D

potential coils for NT x-point

20

21 of 27

DIII-D Neg-D compares well with other no-ELM scenarios

Core:

  • High pressure and confinement due to turbulence reduction
  • Vertical stability is generally worse

Edge:

  • Exhaust integration could be alleviated
  • H-mode and ELMs avoided through MHD?

C. Paz-Soldan, (2021)

PPCF 63 083001

21

22 of 27

Triangularity modifies local shear to close 2nd stability!

  • Local shear defined as:

[A.O. Nelson et al 2022 Nucl. Fusion 62 096020]

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

22

23 of 27

Increased local shear in bad curv. prevents 2nd stability

  • Local shear strong function of geometry�
  • When peak of shear profile ventures into the bad curv. region, access to 2nd stability is lost!

  • This forces a LFS zero-crossing � -> destabilizing

[A.O. Nelson et al 2022 Nucl. Fusion 62 096020]

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

23

24 of 27

For stability, squareness cannot be too high or too low

  • Medium squareness needed for 2nd stability access

  • Still a strong function of 𝛿!

  • Similar to�old DIII-D experiments

[A.O. Nelson et al 2022 Nucl. Fusion 62 096020]

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

24

25 of 27

STEP code used to predict performance on MAST-U

  • STEP allows for integrated modeling of arbitrary devices by connecting utility codes:�
    • CHEF: heating and current drive

    • CHEASE: equilibria

    • TGYRO: transport and profiles� (variable SAT rules)

  • Edge not yet modeled! � (here we force L-mode)

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

25

26 of 27

MAST-U heating contributions chosen to model NBI source

  • Arbitrary heating profiles creating in CHEF to mimic ST NBI�
  • Heating amplitude adjusted to reach convergence with TGYRO

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

26

27 of 27

Difficult to converge TGYRO on stronger NT shapes

rotation fixed

decreasing triangularity

decreasing electron heatflux

fixed heating power

A.O. Nelson / NSTX-U Results Review / PPPL July 25th, 2023

27