The Dialogue Between the Science of Turbulence and Transport and a Burning Plasma Experiment

1
The Dialogue Between the Science of Turbulence
and Transport and a Burning Plasma Experiment

• E.J. Synakowski
• Princeton Plasma Physics Laboratory
• December 11, 2000
• UFA Workshop on Burning Plasma Science
• Austin, Texas

2
There is great value in discussing the scientific
needs of and contributions from a burning plasma
experiment

• Heartfelt, strong sentiments from every vantage
  point
• At issue is the quality of the scientific
  exchange
• What are the issues and concerns a BP must
  address to be viewed as an attractive scientific
  test bed?
• Give take
• What does a burning plasma experiment need from
  the science of turbulence and transport?
• Assume that predicting the performance is of
  value
• What would a burning plasma experiment give to
  the science of turbulence and transport (TT)?
• What the integration goal means to TT
• Flexibility

3
The transport turbulence community speaks
consistently of a couple of major thrusts

• I. A strong desire , what inspires the transport
  community
• scientifically, and what it feels is required to
  advance the field, is to
• develop predictive models based on an
  understanding of
• turbulence and turbulence dynamics.
• From Snowmass
• Goal 1 Comprehensive transport models
• Goal 1a Pursue the challenging, yet realistic
  goal of developing comprehensive predictive
  transport models, based on physically reasonable
  assumptions and well-tested against experiments.
• Goal 1b Detailed Experiment/Theory Comparisons
  at the level of the turbulence

4
The second primary goal develop tools for
turbulence manipulation to control the plasma
pressure

• Again, from Snowmass
• "Goal 2. Develop tools and understanding for
  control of transport and transport barriers...
• This is what the BP language of integration
  speaks to
• Even a doubter of the value of transport barriers
  can benefit from control tools (managing a
  self-heated environment)
• Intrinsically interesting
• Many in the community ask Is a BP the place to
  develop the understanding and the tools?
• Will a BP be able to find a self-consistent
  operating point?

5
What a BP needs from TT issues have been
identified for predictive transport modelling
during the next 10 years

• Highlight items from the Snowmass report are
• Need convergence of models for reliable
  extrapolation
• Reliable models for plasma boundary needed
• Need accurate models for transport barrier
  dynamics (edge and core)

6
There has been progress in developing transport
models that capture experimental trends, but it
is not decisive

• "Are you better off now than you were 8 years
  ago?"
• For temperature profile predictions several
  models are now reasonably successful, but no
  single model is generally preferred. All models
  have flaws and failures.
• The underlying differences between reasonably
  successful models are large enough that we can't
  say that their characteristics point to the
  importance of particular processes
• More complete physics is being included (electron
  dynamics, ExB shear, for example)

7
Turbulence dynamics understanding is key to
developing predictive capability and is a premier
challenge

• Example Zonal flow dynamics
• Several facets have been appreciated only in the
  last several years
• their existence (not verified)
• interaction between zonal flows and
  microinstabilities - saturation levels
• Some consistency in measured features (e.g.
  bursting), but there are other explanations

8
Knowledge of electron thermal transport is key to
predicting BP performance

• Some promising links between theory and
  experiment
• Nevertheless, we are a long way from
  demonstrating that theory explains a lot of
  experimental data.
• Critically important in prediction of a heating
  effectiveness and need for ash removal strategies
• Issue is forcing progress in codes high k modes,
  streamers
• Need for a similar experimental assault

9
Predicting BP performance requires significant
improvement in knowledge of pedestal height and
width

• Many caveats, contradictory theories,
  contradictory
• experiments ? predictive capability not in hand
• Over a range of theories, many have ? r 2/3- 1.
• JT-60U, JET support standard model of ? r
• and gradient near the ideal MHD limit
• Others (DIII-D) support ? independent of r
• perhaps related to second stability
• Progress useful cross-machine database being
  developed
• ITER H-mode Edge Pedestal Expert Group, March
  2000).
• Edge turbulence simulations becoming more
  realistic
• Xu and Cohen (LLNL), Rogers and Drake (U. Md.),
  Scott, Jenko, Zeiler et al. (Garching))

10
Dynamical models capture many features of present
experiments, but cannot yet predict

• Required to assess robustness of operating point
  of a BP and to develop a control strategy
• Self-consistent evolution of turbulence, fluxes,
  shear, and profiles beginning
• Character of some dynamics seen in codes, but
  robust predictions are not in hand

11
Present-day modelling cannot capture dynamics of
new regimes

• For example, an opportunity NSTX
• Lots of exciting speculation about possible
  confinement characteristics - sheared flows,
  aspect-ratio-induced stabilization of drives
  speaks very well of the science
• BUT
• No model has yet dared to try to capture the
  dynamics of the ST in advance in a way that
  guides our experimental choices
• H mode physics, core dynamics, particle
  transport,...
• Lack of predictive capability not intrinsically
  different from the situation on a BP
• Strong self-heating a significant extrapolation.
  Perhaps no stable operating point without
  pressure profile control.

12
The question of what a BP gives back to
turbulence and transport, and the value of the
investment, causes much debate

• If the question is confined narrowly,
• "I have a passion for turbulence and transport
  dynamics what is the best way to learn about
  it?",
• so far, it is not accepted that the BP experiment
  is the experiment
• of choice.

13
The strongest case arises for a BP when two
issues are discussed

• Transport control in a self-heating environment
• But many feel that the flexibility of a D-D
  experiment is better suited for developing
  control tools (e.g. shear flow generation tools)
• r scaling of core transport and edge pedestal
  characteristics
• This is of value in predicting the performance of
  a DEMO or ITER, but the range of r in present
  experiments exceeds the change that a next-step
  BP will provide
• Main value is to prediction of yet another BP
  intrinsic value is debated

14
The BP feature of integration strongly suggests
the need for turbulence manipulation tools to
modify P(r)

• To the Snowmass transport and TTF audiences, the
  BP goal of integration
• speaks to transport barrier control tool
  development.
• Discussions focus on the value of confronting the
  BP, self-heated core and transport control
  self-consistently
• vs.
• for , getting more bang-for-the-buck in one's
  investment by developing a flexible control
  strategy in existing machines
• No doubt the self-consistent alpha heating
  profile will be hard to simulate to everyone's
  satisfaction
• But the science of turbulence manipulation,
  bifurcations, flow shear, Ti/Te effects, would
  benefit greatly from an investment in flexibility
• Of lasting value to a BP is learning to control
  what we have

15
The TT community is motivated by the possibility
of manipulating turbulence dynamics to modify the
pressure profile

• From AT Workshop, GA, March 1999 The single
  issue that was virtually unanimously agreed to be
  the most pressing in terms of the ultimate
  viability of the Advanced Tokamak is the
  following In both the experimental and modeling
  efforts... significant progress... would result
  if local pressure profile control, through
  manipulation of the local transport, can be
  realized...
• The transport community equates the challenge of
  BP integration as the challenge of pressure
  profile control tool development based on
  manipulating turbulence.
• What does a BP contribute to help this effort?
• What development is required in advance of a BP?

16
Flexibility, not rules for enhanced confinement
access, will help ensure the success of a BP
experiment

• The question, "What is the power threshold of a
  core barrier?" is too narrow.
• For example,
• Supershot is transitionless (no bifurcation)
• ERS regime access related to the conditions
  required for Vq (Er) shear layer development
• what is the physics of this event?
• Slow transitions to enhanced confinement
  observed.
• DIII-D NCS, TFTR RS with co-rotation
• Low or no power threshold
• Some transitions with simple q values
• Wide range of dynamics ? BP needs flexible
• tools. There is no scaleable rule.

17
Flexibility allows theories of dynamics to be
challenged

• External Er variations are a powerful teaching
  tool
• Independent variations of Te, Ti push theory to
  different extremes
• The BP community would strengthen their case with
  quantitative discussion regarding flexibility and
  tools, and what will be learned about
• Transport and turbulence dynamics and control,
  e.g. bifurcations, spontaneous shear flow
  generation
• barrier dynamics, expansion, front propagation
• ion vs. electron transport
• that cannot be learned better elsewhere

18
Challenge to the BP community make a scientific
argument for what BP physics can teach about
transport and turbulence dynamics

• The TT community is passionate about
  understanding of turbulence and turbulence
  dynamics
• The scientific basis for predictive capability
• Intrinsically front-line physics, and a great
  export to other fields
• Flexibility, accessibility essential for making
  progress in a BP or elsewhere
• Advances being made, but ability to predict
  performance of a BP has not changed in a decisive
  sense
• Flexibility is required for the integration goal
  of a BP to be compelling to TT
• Tools to vary Er, Ti/Te, diagnostics need high
  priority
• Control tool development required to ensure a
  stable operating point
• Many feel that the place to do this is in an
  invigorated base program
• Integration ltgt P(r) control via turbulence
  manipulation