Working Group IV

National Spherical Torus Experiment

E. Synakowski (Coordinator, Princeton Plasma Physics Laboratory)

D. Newman (Assistant Coordinator, Oak Ridge National Laboratory)

M. Beer (Princeton Plasma Physics Laboratory)

W. Dorland (University of Texas at Austin)

R. Fonck (University of Wisconsin)

C. Greenfield (General Atomics)

E. Mazzucato (Princeton Plasma Physics Laboratory)

T. Peebles (University of California at Los Angeles)

G. Rewoldt (Princeton Plasma Physics Laboratory)

G. Staebler (General Atomics)

M. Zarnstorff (Princeton Plasma Physics Laboratory)

The National Spherical Torus Experiment (NSTX) presents many unique scientific challenges from the point of view of confinement studies and understanding of local transport. This report represents an effort to outline key physics issues pertaining to global confinement, local transport, and fluctuations that should be addressed by the NSTX research program in the first two to three years of operation. The issues were discussed at the NSTX Research Forum, held on February 5 - 7, 1997, at the Princeton Plasma Physics Laboratory. As a result of these discussions, recommendations have come forth with respect to a research program outline and identification of work needed in the near term with respect to diagnostic development and theoretical modeling. The last item is not a purely academic exercise: many of these modeling efforts are needed to help experimentalists make judgments for diagnostic systems based on a consistent set of underlying assumptions.

This report is organized as follows. First, an outline of the major physics themes pertaining to confinement is given. Second, issues pertaining to diagnostics are outlined, both with respect to the schedule of bringing these items on-line, and with respect to technical issues that must be addressed aggressively in order to meet the physics goals. Third, theory-related issues are highlighted, most of which bear either on diagnostic development in the near term or on influencing early experimental programmatic choices.

It is recommended that the near-term (2 - 3 year) confinement studies program on NSTX first focus on issues for which aspect ratio plays a particularly strong role in the underlying physics. Highest priority is given to those topics for which this characteristic might be used to illuminate underlying commonalities with normal aspect ratio tokamaks.

A. Contribute to the confinement scaling database. This issue will likely be performed in the early days of operation of NSTX, when diagnostics are just coming on-line. Global confinement scaling with average electron density <ne>, plasma current, RF heating power, and aspect ratio would be explored.

Further efforts in confinement scaling studies should quickly move from the global to the local, including characterization of thermal, momentum, and particle transport coefficients from power and particle balance considerations. Such a campaign will demand more extensive profile coverage than the "Day 1" diagnostic set lists. This issue is highlighted by the simplest consideration of the desired programmatic elements, and is revisited in the discussion of the individual measurements that are needed (Section III).

B. Aspect ratio scaling. Results from past scaling studies suggest a dependence of confinement on toroidicity (strong dependence on major radius, weaker dependence on minor radius [Grisham et al., Phys Rev. Lett. 67 (1991) 66]) that serves as part of the motivation to perform confinement studies that focus on variations in aspect ratio. A second motivation is born from the work of Rewoldt et al. [Phys. Plasmas 3 (1996) 1667], where it has been suggested that partial or complete stabilization of certain microinstabilities might be expected for A < 1.5. Results from such a study will serve to illuminate issues basic to the viability of an ST as a reactor concept. Again, global studies can be performed initially, but rapidly the scientific need for profile diagnostics, including core fluctuation measurements, will become urgent.

The most relevant aspect ratio studies will be those in which other parameters important to core confinement are kept constant. Pertaining to this, we note that for scans in which neutral beam heating is applied, there is an apparent difficulty in keeping the ratio of the heating profile width to the plasma height constant as A is varied. This is a fundamental limitation owing to the height of the TFTR heating beams. The possibility of introducing an insertable mask at the grids of the neutral beam ion source might be investigated as a means of altering this height, and keeping the heating profile self-similar during such a scan. Of course, RF-only aspect ratio scans could be performed, and with sufficiently narrow deposition profiles would not be compromised by an analogous issue.

Two emphases are suggested. First, the dimensionless scaling studies should be designed to make contact with the results obtained on tokamaks with more typical aspect ratio. It should be determined whether it is possible to obtain matches in dimensionless parameters expected on NSTX on other devices. The possibility of enlisting the support of other tokamak devices, such as DIII-D, should be explored. This support might come in the form of experiments in which the expected dimensionless parameters of NSTX are matched as closely as possible. Second, connection should be made with other small aspect ratio devices. It is noted that the Pegasus device (University of Wisconsin) will connect with the available aspect ratios for NSTX (1.25 ~ 2) from the lower end, making a comparison of confinement scaling results (in any sense, not just dimensionless scaling) from that device to NSTX important.

D. H mode transitions: L-H mode transitions and the edge pedestal width. Predictions of the performance of small aspect ratio devices include the possibility that the power threshold for H modes will be quite low on NSTX. The wide ranging possibilities regarding this phenomena include the prospect that transitions to enhanced confinement might not be observed at all, and that the normal state of operation might always be one of strong suppression of edge turbulence without the onset of an edge barrier. The speculations regarding this issue are wide-ranging. NSTX power threshold studies will complement those performed on Alcator C-MOD, as both devices have small major radius compared to most of the contributors to this issue, but have B fields that differ by as much as two orders of magnitude at the edge.

E. Core transport barrier studies. It has been suggested that small aspect ratio devices might easily promote the onset of reduced transport by the formation of core transport barriers [Stambaugh, IAEA 1996, Montreal; T.S. Hahm, synopsis contribution to the Forum] Contributing factors include the expected reduction in growth rates at low A, and the large pressure gradients permitted by the high beta limits that are expected in ST's.

F. Perturbative transport, and comparison to steady-state properties. One of the most powerful means of revealing the underlying dependencies of particle and thermal fluxes is through the application of perturbations, and the comparison of perturbed fluxes to those inferred from steady-state particle and power balance techniques. Such perturbations include working gas and impurity gas puffing, pellet injection, neutral beam blips, and the application modulated or pulsed RF. Such studies are particularly relevant to NSTX, as a result of the unknown scaling of fluctuation behavior to small A, and the predictions that instabilities dominant in normal aspect ratio devices such as TFTR ought to be stabilized at the smallest A values. As such, the response of fluxes to perturbations ought to illuminate any fundamental changes between normal and small A. These perturbation studies should be part of any of the aspect ratio and dimensionless scaling programs on NSTX.

The relation between particle and energy transport should also be a focus of both perturbative and steady-state transport analysis. In particular, the viability of any reactor concept hinges in part on the relationship between impurity (helium) transport and thermal flows. Helium transport studies, performed with spectroscopy with gas puffing and neutral beam injection, can be used to assess this important issue.

A significant concern of the Working Group pertained to the planned implementation of profile diagnostics. As such, two strong conclusions were reached:

It is recognized that measurements of some quantities, such as the current profile, requires careful consideration and development.

With the refocused emphasis of the fusion effort as a science-based plasma physics program, and with the opportunity to use the spherical torus concept as a test bed for obtaining leverage with respect to many fundamentally important physics issues, it is strongly felt that developing the means of measuring local thermodynamic quantities early in the NSTX program should be made a high priority. Given the vigor and increasing sophistication of the dialogue between experiment and theory, and the possibly profound physics issues that may be illuminated by examination of the ST concept, a physics-based national program of ST research cannot survive long without both temporal and radial documentation of quantities such as electron and ion temperature, flow velocities, and magnetic field.

Issues pertaining to the measurement of various quantities, and possible diagnostic development issues and recommended tasks, are highlighted below.

A. Electron temperature measurements. Time-dependent measurements of electron temperature are of high value on both TFTR and DIII-D. Thomson scattering is the only electron temperature diagnostic which is guaranteed to work for NSTX. The value of multiple time point radial coverage with Thomson scattering systems has been demonstrated in the community on the DIII-D device. The alternative of a single time-point system is deemed unacceptable when considering the exploratory nature of early plasma operation on any device, and the potential importance of performing studies of time dependent and non-linear behavior on NSTX. In addition, the diagnostic cost of implementing of a multiple time point system has to be weighed against the cost of compiling time histories from multiple shots, and the possible physics studies that might be compromised if the program is forced to rely on a single time point system.

Such systems have proven their worth not only in terms of temperature measurements, but density measurements as well. This may be particularly important in assessing the behavior of the edge region in H mode plasmas. The value of configuring multipoint systems in "burst" modes in order to characterize plasma density and temperature behavior on fast time scales will also be of value in the study of nonlinear plasma phenomena.

It is also proposed that the technique of using the Electron Bernstein Wave emission for electron temperature and electron temperature fluctuation measurements be explored. This promising technique must be examined from the point of view of wave propagation physics and launch. A possible test bed for this technique might be the CDX-U device at PPPL.

Use of soft x-ray measurements to get the central electron temperature and a measure of the profile shape through filtered soft x-ray arrays is also suggested. The arrays may also be useful for MHD instability identification, as well as a measure of the equilibrium flux surface shape as demonstrated on PBX-M [[E.T. Powell, R. Kaita, and R.J. Fonck, Rev. Sci. Instrum. 61 (1990) 3301]

Recommendations for near term action:

B. Ion temperature and ion rotation measurements. Assessing local transport on any auxiliary heated plasma device requires separation of the roles of the ion and electron channels. Charge exchange recombination spectroscopy (CHERS) has proven to be invaluable throughout the community as a means for measuring ion temperature, impurity ion density profiles and fluxes, and plasma rotation. In addition, CHERS measurements provide many of the elements required to measure the radial electric field through solution of the force balance equation for the observed impurity. This will be of central importance to NSTX in the study of transport barrier formation and sustainment. Both of these issues underscore the desirability of having the TFTR heating beam installed as quickly as possible. New high throughput transmission grating spectrometers may be considered for this measurement system.

As a result of relatively cool edge temperatures and the proximity of the sight line paths to impurity sources, it is quite likely that edge emission of hydrogen-like impurities will compete with charge exchange emission from the neutral beam volume. Given the relatively good diagnostic access on NSTX, it is urged that a viewing location be allocated to CHERS to install a fiber optic array that has a similar orientation as that viewing the neutral beam, but does not view the neutral beam directly. This will enable a direct measure of background light, a problem that plagues CHERS measurements on every toroidal device where they are attempted. Access to a vertical view of the heating beam also is not available as the vacuum vessel is presently envisioned. It is strongly recommended that modifications be made soon to allow for future flexibility, even if it is decided that the installation of vertically view optics should be delayed.

Note also that there is concern over the interaction of fast ions injected by neutral beams and RF applications on NSTX. This concern underscores the potential value of an X-ray crystal spectrometer to measure ion temperature in the event that neutral beam heating is not desired for a particular experiment (or before the beam is available) , such as one employing RF heating.

Recommendations for near term action:

C. Electron density measurements. While a multiple time point Thomson scattering system might be used, interferometry must seriously be considered for this needed measurement due to its superior time response. The existing TFTR MIRI system is a candidate for this measurement. However, it is pointed out that while vertically oriented chordal measurements provide the possibility of being combined with Faraday rotation measurements of the current density, their inversion requires knowledge of the flux surface geometry. The inversions of interferometry data might be aided by constraints on the flux surfaces provided by x-ray imaging (Powell et al.; see section D below).

Recommendations for near-term action:

It is also suggested that measurement of the pitch angle of pellet ablation clouds be considered for q profile measurements in the core. However, the required pellet injection would of course destroy the plasma properties of interest. Finally, equilibrium reconstruction aided by measurements of flux surface geometry via pinhole camera images has been demonstrated on PBX-M [E.T. Powell, R. Kaita, and R.J. Fonck, Rev. Sci. Instrum. 61 (1990) 3301]. The viability of this technique should be assessed for NSTX plasmas.

Recommendations for near term action:

E. Core fluctuations. Predictions of the suppression of turbulence as low aspect ratios are approached [Rewoldt, '96] underscores the strong desirability on scientific grounds of measuring characteristic core fluctuations on NSTX. In existing devices, dialogue with the theory community is becoming increasingly dependent on such measurements.

In the expected parameter range in NSTX electron cyclotron emission-based scattering techniques will not be possible. O-Mode reflectometry in the 20 - 60 GHz range may provide a means of making core measurements, as it is well suited to the small poloidal wave numbers expected. This technique's utility is limited to density profiles with some degree of peaking. Also, launch issues pertaining to the pitch angle of the magnetic field need to be assessed. In the event that flat density profiles are dominant on NSTX, collective FIR scattering can be performed, but little spatial and k resolution will be possible.

A potential difficulty of fluctuation measurements by Beam Emission Spectroscopy (BES) was identified in this workshop, namely, the presence of large poloidal pitch angles, especially near the plasma edge. On TFTR and DIII-D, advantage is taken of the fact that midplane sightlines view along a field line within the neutral beam volume, thus providing poloidal spatial resolution. On NSTX, such a midplane view would result in poor resolution, since the pitch angle of the field lines will be many tens of degrees. Core measurements with BES appear to be more tractable using midplane views.

Note that, although long wavelength modes are expected, the program should be prepared to examine the shorter wavelength regions as well. This issue may be of particular importance to the understanding of the anomalously high electron transport observed on larger aspect ratio devices even when the ion channel is suppressed. The essential point is that, while informed speculation about fluctuation properties exists, ST's represent such a radical departure from existing experiments that the program should be prepared for deviations from expectations, which will undoubtedly occur.

In the Working Group discussions, it became evident that more discussion is required to come to a recommendation regarding the best approach for core fluctuation measurements. Further discussion, with the possible inclusion of others experienced in fluctuation measurements in the plasma core, seems in order.

Recommendations for near term action:

The implementation of a heavy ion beam probe should also be considered for this application.

Recommendations for near term action:

G. Edge transport diagnostics. The topic of fluctuations and transport is of utmost importance at the plasma edge as well. As Langmuir probes are relatively inexpensive, and the edge of ST devices is unexplored territory, they should be implemented in such a way to permit poloidal coverage as well as midplane coverage. Reciprocating probes should also be considered. As mentioned above, reflectometry techniques should also be considered for edge electron density measurements. They also can provide poloidal coverage.

Multilayer mirror spectroscopy is also suggested as a possible means for obtaining edge temperatures and impurity information, and its potential for characterizing edge MHD activity has been demonstrated.

Discussion of diagnostic and other programmatic issues would be greatly facilitated by the widespread use of standardized profiles.

Forward modeling with a self-consistent equilibrium, under some agreed-upon set of assumptions, should be performed to generated standardized profiles of electron density, electron and ion temperature, impurity concentration, q profile, and toroidal rotation profile. Three sets of conditions were identified as being potentially useful:

While some profiles do exist that have been optimized with respect the to beta limits and the plasma pressure, they are not widely available in general. Also, since the future performance of NSTX is such an unknown, the availability standardized profiles over a wide range of confinement possibilities is highly desirable.

Examination of a number of theory-related issues should be pursued in the near term.

A. Kelvin-Helmholtz instabilities. Due to the presently envisioned unbalanced beam injection, attention should be paid to the possible onset of Kelvin-Helmholtz instabilities for NSTX-like plasmas in gyrokinetic simulations. This can be examined as a matter of course in gyrokinetic studies of microinstabilities in general, but only if rotation is explicitly included in the assumed profiles. This again points to the need for a self-consistent estimate of a range of obtainable toroidal rotation velocities. The issue is potentially of concern for NSTX, owing to the stronger toroidicity of small aspect ratio plasmas.

B. Neoclassical performance of NSTX plasmas. A related issue was pointed out above, but the importance of understanding the implications of neoclassical transport for NSTX is emphasized here by pointing out that there may be issues for MHD stability if somehow neoclassical transport is realized. Is neoclassical transport, and the accompanying pressure gradients, consistent with the stability requirements of NSTX? What is the beta limit with the available power if neoclassical theory dominates?

Related to this issue is an assessment of the applicability of neoclassical theory to ST's in general. Does neoclassical theory exist in sufficient sophistication to cope with the large orbit sizes that will occur in ST's, especially in light of the possibly steep pressure and electric field gradients? Will the neoclassical transport really be small, given the large orbits

C. Nonlinear neoclassical viscosity regimes. Strong damping of poloidal flows is expected in NSTX, due to the large gradient in the toroidal magnetic field. However, due to large orbit and drift orbit widths and the expected high beta limit, significant orbit squeezing is expected. Do these effects lead to the possibility of approaching a nonlinear neoclassical viscosity regime in which the damping is overcome?

D. Microinstability and nonlinear analysis of standardized profiles. The standardized profiles discussed above should be checked with microinstability calculations and nonlinear simulations to determine whether they are reasonably self-consistent. Preliminary work in this area has begun [Rewoldt], but the profiles should be varied in a way consistent with the expected input power.