Experimental Diagnostics

Diagnostic Improvements

A dramatic improvement in the understanding of the turbulent mechanisms driving anomalously high levels of cross-field energy and particle transport in magnetic confinement devices has been achieved during the past decade. This accomplishment is due in large part to the combination of theoretical simulations of turbulence, detailed profile measurements of density and temperature, and the development of fluctuation diagnostics to experimentally characterize the underlying turbulence. Nevertheless, a great deal remains to be learned and the development of more advanced fluctuation diagnostics, improvements in existing diagnostics and implementation of profile and fluctuation diagnostics on a greater variety of magnetic configurations will play a crucial role in increasing our understanding of the basic physical principles, as well as provide greater confidence in our predictive capabilities.

Particular issues to be addressed include:

Turbulent mode identification
LH transition physics
ExB shear dynamics
Broad ranging issues including indications of the universality or self-organizing nature of plasma turbulence across various magnetic configurations
The potential for stabilization in ST and ET devices.

It is desired to obtain fluctuation measurements of several plasma quantities (n, T_i , T_e , f , B, v) over a wide range of wavelengths (0.5 < k < 50 cm ^-1 ) to most fully characterize turbulent driven transport (G=<nv_r >, q=<nvT>, ...). Existing fluctuation diagnostics are focussed dominantly, though not exclusively, on the turbulent density field and development has emphasized application to tokamaks.

Examples of such existing diagnostics include:

scattering (microwave, far-infrared, CO 2 ), (correlation)
reflectometry (local ñ, L c,r (radial correlation length))
beam emission spectroscopy (local ñ, L_c,r L_c,f )
Langmuir probes (local ñ,~f at edge/SOL)
phase contrast imaging (ñ, S(k r ))
among others

 

In addition, electrostatic potential fluctuations are measured with Heavy-Ion Beam Probes, through with limited application. Electron temperature fluctuations can in principle be obtained with correlation ECE, though this technique is under development. Limited ion temperature fluctuation measurements have been obtained using fast charge exchange recombination spectroscopy. Magnetic fluctuations are crucial to understanding RFP (and perhaps other) transport, and possibilities for measuring this include cross-polarization scattering, though further demonstration is required. These techniques have made valuable contributions, yet can all benefit from improvements (e.g., temporal and spatial resolution, signal-to-noise ratio). It is also noted that while the primary focus of fluctuation diagnostics is for transport and turbulence studies, these diagnostics have been applied to examination of MHD phenomena (e.g., TAE, NTM mode structure), and to measurements of RF mode-coupling behavior, and so have a broader application within MFE research.

There is significant interest in obtaining two dimensional images of turbulence. Turbulence in magnetically confined plasmas is fundamentally quasi-2D (3D but very elongated along the magnetic field) in nature, and simulations can now model the 3D nonlinear dynamics, while most fluctuation measurements are 1D or point measurements. It is therefore desired to provide comparable 2D measurements for visualization applications as well as to more fully examine the nonlinear mode coupling features.

Some concepts under development or being considered for 2D imaging of density fluctuations include:

edge/SOL Langmuir probe arrays
imaging BES
laser induced fluorescence
reflectometry
fluorescent plate imaging (cool plasmas)....

 

Another key area in which to develop advanced diagnostics is higher-k measurements. While ion transport appears to be fairly well understood in tokamaks, knowledge of electron particle and energy transport remains far more elusive. An area of intense theoretical and experimental study is higher-k modes (e.g., ETG) that may be controlling electron thermal transport in some regimes. As such, it is desired to develop high-k (k > 5 cm ^-1 ) diagnostics for any and all plasma quantities.

Possibilities include:

high-k FIR scattering techniques

 

Profile and fluctuation measurements are fairly well developed on the tokamak, yet are required for other configurations. The existing methods often suffer from either not being applicable to other configurations, or simply being too expensive to deploy on smaller concept development machines. Efforts should therefore be made to provide profile diagnostics, required for transport studies, as cheaply as possible. This might be accomplished by developing less expensive profile measurement techniques, as well as by utilizing equipment from "retired" machines, if applicable, to the maximum extent possible. Other options may include more limited profile diagnostics: fewer channels, lower time resolution. Similar issues apply to implementing fluctuation measurements on such devices. Table 1 summarizes a list of existing profile and fluctuation diagnostics and where they have been or could reasonably be deployed.

We also note that it is desired to implement "simulated diagnostics" in the various modeling codes so that nonlinear simulation results can be directly compared with fluctuation measurements. This would involve applying the finite spatial resolution and volume sampling, time resolution, and providing lab-frame equivalent measurements (incorporation of plasma rotation, radial electric field effects). This topic is addressed in more detail in the experiment/theory comparison section. Through a combination of applying existing diagnostics (profile and fluctuation) to non-tokamak magnetic configurations, further optimizing existing diagnostics, and development of new diagnostics for as yet unmeasured quantities, and direct comparison with turbulence simulations, a more thorough understanding of the fundamental nature of plasma turbulence will be obtained.

TABLE 1. Density fluctuation measurements on existing and developing configurations

Diag.	Tokamak	Sph. Tok.	Stellarator	RFP	Spheromak
Scattering	X	A	X	A	A
Reflectometry	X	A/D	X	A/D	-
BES(beam)	X	D	A	-	-
Probes	X	X	X	X	X
PCI	X	A	A	A	A
Thomson scat	X	A	X	X	A
CER(beam)	X	A	X	A	A
ECE	X	-	X	-	-

X-Experimentally demonstrated, A-Applicable, D-Difficult, - Doesn't apply

Improve Experiment/Theory/Computational Comparison and Coordination

In order to further both the scientific goals of understanding turbulence and transport in plasmas and the programmatic goal of utilizing that understanding to control transport and make fusion a practical energy source, an improved comparison between experiment, theory and computation is needed. In order to achieve this goal, improvements (or changes) are needed in all three areas. In the experimental area, dedicated run time must be allocated for experiments specifically designed to test aspects of the theory using relevant diagnostics. This often will mean simply an improved utilization of existing diagnostics in operating regimes thought to overlap with modeling and theory capabilities. However, in some cases, development of novel new diagnostics will be required to make measurements that can be compared with theory and computation (e.g. localized short wavelength fluctuation measurements for comparison with ETG and other theory predictions). In the computational area, simulated diagnostics should be "built" for more realistic comparison with experimental measurements. The models should also be run for as wide a range of "reasonable" parameters (that is, parameters which are both consistent with the assumptions built into the model and in physically realizable regimes) as possible. In the area of theory, experimentally verifiable predictions should, when possible, be a goal. Coupled with the individual efforts, enhanced analysis techniques are needed to facilitate the actual comparisons. This is because all measures are not equal in comparisons (e.g. different. models can give the same fluctuation spectra and fluctuation mean etc). Very often it is in the tails of the probability distribution functions (PDFs), the parts most difficult to investigate, that the differences show up. Therefore, a hierarchy of comparison techniques should be employed.

These can include;

Basic statistics (mean variance etc)
Spectra (frequency and k, with long enough saturation regime)
Time evolving spectra (plasma turbulence spectra are rarely steady stationary, wavelets can be useful)
Bi-spectral techniques to infer growth rates (both linear and nonlinear) nonlinear transfers etc
Quantile-quantile (for non-normal comparisons)
Correlation functions (auto and cross, spatial and local)
Structure functions (to investigate intermittency etc)
R/S analysis (to investigate system dynamics).

All of these techniques can and should also be used for inter-model comparisons, inter-experiment comparisons (to help address the issue crosscuts) and innovative concept experiments and true predictive theory-model comparisons as well as computation-experiment comparisons. Successful comparisons will form a basis for predictions of new regimes, leading to next step capabilities.