Measurement of Turbulence Decorrelation during Transport Barrier Evolution in a High-Temperature Fusion Plasma

A low power polychromatic beam of microwaves is used to diagnose the behavior of turbulent fluctuations in the core of the JT-60U tokamak during the evolution of the internal transport barrier. A continuous reduction in the size of turbulent structures is observed concomitant with the reduction of the density scale length during the evolution of the internal transport barrier. The density correlation length decreases to the order of the ion gyroradius, in contrast with the much longer scale lengths observed earlier in the discharge, while the density fluctuation level remain similar to the level before transport barrier formation.

Figure 1

(a) Full wave simulation of the intensity of launched and reflected microwaves from JT-60U plasma. Density fluctuations of 0.5% are imposed on the plasma reflecting layer. An enlargement of the reflecting region is shown in (b).

Figure 2

The evolution of plasma parameters during the formation of the internal transport barrier: (a) plasma current and neutral beam heating power; (b) ion temperature at two radii; and (c) evolution of the density profile. The red arrow in (c) corresponds to the trajectory of the $X$-mode reflecting layer at 115 GHz. Toroidal field 3.5 T, plasma current 1.3 MA, major radius 3.2 m, minor radius 0.96 m.

Figure 3

(a) Spectrum of reflected waves, (b) correlation vs radial separation of reflecting layers, and (c) radial correlation length vs the density scale length in the transport barrier.

Figure 4

Radial correlation data and full wave simulations at 2 times, before high power neutral beam heating in (a) and in the fully developed internal transport barrier in (b). The measured correlations (red squares) are compared to 1D (solid line) and 2D full wave simulations (diamonds). In (a) a density correlation length ${\lambda }_r\simeq 20\mathrm{cm}$ ($k_r=0.1{\mathrm{cm}}^{-1}$) is inferred consistent with the 1D and 2D modeling. In (b) data are shown in red squares, and 1D modeling results are shown for $k_r=5{\mathrm{cm}}^{-1}$ (solid line) and $8{\mathrm{cm}}^{-1}$ (dashed line). The 2D modeling shows that the estimated correlation length is only weakly dependent on the choice of the poloidal (vertical) scale length.

Figure 5

Inferred density fluctuation level (blue) and the plasma density profile (red) in the target plasma (a) and in the internal transport barrier plasma (b) compared to the calculated linear growth rate (bottom) using the FULL code without rotation (purple) and with rotation (green). The calculated linear growth rate is only plotted in the region where there are reflectometer measurements.