Page 6 - Robustness of waves with a high phase velocity

FIGURE 4 Time evolution of relative density fluctuation level vs. radius (plasma frame); (a) plasma gun not activated; (b) plasma gun active.
E×B SHEAR AND TURBULENCE DECORRELATION
We now examine the spatial structure of the E×B shear and its influence on the turbulence level. Figure 5(a,d) show radial profiles of the radial electric field, evaluated from the measured E×B velocity. The radial profile of the axial magnetic field has been estimated from the FRC rigid rotor model [31], approximating the measured density profile by a rigid rotor profile. The electric field is positive outside the FRC separatrix and reverses in the FRC core. It has been predicted that turbulent eddies are elongated and eventually sheared apart if the flow shearing rate exceeds the turbulence decorrelation rate [33], a process shown to suppress edge turbulence in the low-to high confinement-mode transition (L- to H-mode transition), and in internal barriers in tokamaks [34,35]. The shearing rate is determined from the differential flow between two neighboring DBS probing locations r1,r2 . The turbulence decorrelation rate is estimated from the width of the auto-correlation coefficient of the backscattered signal.
Fig. 5(b,c) shows a comparison of the E×B shearing rate with the turbulence decorrelation rate ΔωD. With E×B turbulence advection present, the decorrelation rate in the laboratory frame is given by The radial turbulence correlation length λr is measured by DBS (λr~2-6 cm here, which is on the order of the ion gyroradius). Estimating the poloidal correlation length ≥ λr, we obtain and to good approximation ΔωD ~ ΔωDlab.
Importantly, when the plasma gun is not active, is observed to exceed the decorrelation rate early in the shot (220 μs) over a ~0.1m wide radial range, but at a later time (340 μs). At that time, fluctuation levels and radial transport increase and FRC radius/flux confinement begin to decline rapidly [Fig. 2(a,b)].
The toroidal wavenumber range, determined by the beam launch angle ζ and the plasma geometry, is centered around ~ 2-3 cm-1 ( ~ 2), with ∼1.8−2.3 due to plasma curvature [25] for the data in Fig. 3; hence a relatively broad wavenumber range is included in the measurement.
In contrast, in shot #20957 (plasma gun active), the shearing rate exceeds the turbulence decorrelation rate early (280 μs) as well as later (450 μs) in the discharge [Fig.5(e,f)]. Correspondingly, magnetic flux confinement is improved (the FRC separatrix does not contract rapidly as in shot #19896 without plasma gun [Fig.3 (a,b)]), and radial particle and thermal transport is reduced. The n=2 mode is absent, possibly due a combination of axial line- tying due to the gun-injected SOL plasma, and/or modification of the E×B and toroidal ion rotation velocity.
CONCLUSIONS
In conclusion, we have characterized density fluctuations with toroidal scale lengths on the order of and smaller than the ion gyroradius in a large FRC plasma. Fluctuation levels peak in amplitude in the region of strong radial density/temperature gradients just outside the FRC separatrix. Density fluctuation levels ñ/n near the separatrix and in the SOL increase beyond a critical density gradient. The observed strong SOL turbulence is ascribed to drift- or drift-interchange modes driven unstable by the radial density and/or temperature gradients, possibly in combination with unfavorable curvature in the open field line, mirror-confined SOL plasma region. Further experimental and modeling/simulation work is under way to identify unambiguously the underlying instability drive [36].
030002-6