A new system of gyro-Landau fluid (GLF) equations for tokamak plasmas is presented. The new equations include both trapped particles, which can average the Landau resonance, and passing particles which do have a Landau resonance. The trap GLF (TGLF) model is unrestricted in trapped fraction or perpendicular wave number of the electrostatic perturbation. The linearly unstable eigenmodes of the TGLF equations include low-frequency trapped ion modes all the way up to high-frequency electron temperature gradient driftwaves. Extensive benchmarking of the linear TGLF eigenmodes with a large database of gyrokinetic linear stability calculations verifies that the TGLF model is accurate over the full range of plasma parameters tested.

The effects of E ϫ B velocity shear have been investigated in nonliner gyrokinetic turbulence simulations with and without kinetic electrons. The impact of E ϫ B shear stabilization in electrostatic flux-tube simulations is well modeled by a simple quench rule with the turbulent diffusivity scaling like 1 − ␣ E ␥ E / ␥ max , where ␥ E is the E ϫ B shear rate, ␥ max is maximum linear growth rate without E ϫ B shear, and ␣ E is a multiplier. The quench rule was originally deduced from adiabatic electron ion temperature gradient ͑ITG͒ simulations where it was found that ␣ E Ϸ 1. The results presented in this paper show that the quench rule also applies in the presence of kinetic electrons for long-wavelength transport down to the ion gyroradius scale. Without parallel velocity shear, the electron and ion transport is quenched near ␥ E / ␥ max Ϸ 2 ͑␣ E Ϸ 1/2͒. When the destabilizing effect of parallel velocity shear is included in the simulations, consistent with purely toroidal rotation, the transport may not be completely quenched by any level of E ϫ B shear because the Kelvin-Helmholtz drive increases ␥ max faster than ␥ E increases. Both ITG turbulence with added trapped electron drive and electron-directed and curvature-driven trapped electron mode turbulence are considered.

Plasma turbulence characteristics, including radial correlation lengths, decorrelation times, amplitude profile and flow velocity, have been measured during a ρ * scan on DIII-D while all other transport relevant dimensionless quantities (e.g., β, ν * , κ, q, Te/Ti) are held nearly constant. The turbulence is measured by examining the correlation properties of the local long wavelength (k ⊥ ρi ≤ 1) density fluctuations, measured with beam emission spectroscopy. The radial correlation length of the turbulence Lc,r is shown to scale with the local ion gyroradius, Lc,r ≈ 5ρi, while the decorrelation times scale with the local acoustic velocity as τc ∼ a/cs. The turbulent diffusivity parameter, D ∼ (Lc, r 2 /τc), scales in a roughly gyro-Bohm-like fashion, as predicted by the gyrokinetic equations governing turbulent transport. The experimental one fluid power balance heat diffusivity scaling and that from GLF23 modelling compare reasonably well.

The shear in the mean field velocity Doppler shift is shown to suppress the amplitude of electric potential fluctuations by inducing a shift in the peak of the radial wave number spectrum. An analytic model of the process shows that the fluctuation spectrum shifts in the direction where the velocity shear is linearly destabilizing but that nonlinear mixing causes a recentering of the spectrum about a shifted radial wave number at reduced amplitude A model for the 2D nonlinear spectrum is used in a quasilinear calculation of the transport that is shown to accurately reproduce the suppression of energy and particle transport and the Reynolds stress due to the velocity shear.

This paper reports on over 100 nonlinear simulations used to systematically study the effects of safety factor q and magnetic shear ŝ on turbulent energy and particle transport due to ion temperature gradient (ITG) modes and trapped electron modes (TEM) for several reference cases using the GYRO gyrokinetic code. All the simulations are collisionless, electrostatic, and utilize shifted circle geometry. The motivation is to create a database for benchmarking and testing of turbulent transport models. In simulations varying q, it is found that the ion and electron energy transport exhibit an offset linear dependence on q for 1⩽q⩽4. This result is valid for cases in which the spectrum is dominated by either TEM or ITG modes. The particle transport also follows a linear q dependence if the diffusivity D is positive (outward). If a particle pinch is predicted, however, then D is found to be insensitive to q. In kinetic electron simulations varying the magnetic shear ŝ, the particle transport can exhibit a null flow at a particular value of ŝ. In the vicinity of the null flow point, the transport spectrum shows that some modes drive an inward flow while others drive an outward flow. For negative magnetic shear, the magnetohydrodynamic α parameter is shown to be stabilizing for both the energy and particle transport but can be destabilizing for large positive shear. Compared to the ITG dominated case, the TEM cases show the same linear q dependence, but a weaker ŝ dependence is exhibited for positive magnetic shear values when TEM modes dominate the spectrum. In general, the q, ŝ, and α dependence of the transport including kinetic electrons is consistent with ITG adiabatic electron simulation results.