Geometry dependence of stellarator turbulence

Spatial structure of drift-wave turbulence and transport in a stellarator

“…Therefore, the progress in the magnetic confinement of plasmas is inevitably related to our understanding of driftwave turbulence. In addition to turbulence driving or reducing background parameters like pressure gradients and radial electric fields [3], the magnetic field geometry is able to modify the spatio-temporal appearance of drift-wave turbulence [4]. Local parameters of the magnetic field can generate an additional drive of instability and lead to spatial inhomogeneities of turbulent transport.…”

“…Especially in complex stellarator configurations, the local properties of the magnetic field, like curvatures and field-line shear, have to be included into the models in order to describe the structure and dynamics of turbulence appropriately. Different approaches like a ballooning-mode formalism [6,7,8,9,10,11], a full three-dimensional analysis of unstable modes on a flux surface [12,13,14,15] and non-linear fluid [16,17,18] or gyrokinetic [19,4,20,21] simulations point to significant effects of the local magnetic field geometry on the spatial structure and the stability of drift modes. Theoretical studies indicate that turbulent transport can be reduced by more than a factor of 2 if an appropriate shaping procedure considering local magnetic field properties is applied [22].…”

Gyrokinetic Vlasov Code Including Full Three-dimensional Geometry of Experiments

“…For more quantitative gyrokinetic simulations, it is a natural path to furnish a well-established gyrokinetic code with detailed geometrical information obtained from three-dimensional equilibrium calculations as in Refs. [15][16][17]. Based on this motivation, we developed a new gyrokinetic Vlasov code, GKV-X.…”

Optimizing Stellarators for Turbulent Transport

“…To surmount this obstacle, we instead employ a "proxy function" Q prox in C 2 t to stand in place of Q gk , a fairly simple function of key input geometric quantities, based on theory and on the geometry dependences of Q gk found in Gene studies on a family of nc-optimized stellarators. [9] Q prox need not give a highly accurate prediction of what the gk result will be (though of course the more accurate the better) -it need only capture enough of the physics to guide the optimizer toward configurations which Gene will subsequently confirm has reduced Q gk . Moreover, by examining the means by which Stellopt contrives to improve Q prox and Q gk , one may learn methods for deforming the stellarator shape to achieve the turbulent stabilization which are geometrically possible, whose discovery without the optimizer would be extremely difficult.…”

“…9, for simplicity we consider only ion temperature gradient (ITG) turbulence [10] with adiabatic electrons. As found there, two geometric quantities central to determining the form and amplitude of the turbulence are the "radial curvature" κ 1 ≡ e x · κ, with vector curvature κ and e x the covariant basis vector for x, [8] and the local shear s l ≡ ∂ θ (g xy /g xx ), with θ the poloidal azimuth in flux coordinates, which parametrizes distance along a field line.…”