Abstract
We present a convergence study of the gyromoment (GM) approach, which is based on projecting the gyrokinetic distribution function onto a Hermite–Laguerre polynomial basis, focused on the cyclone base case (CBC) (Lin et al., Phys. Rev. Lett., vol. 83, no. 18, 1999, pp. 3645–3648) and Dimits shift (Dimits et al., Phys. Plasmas, vol. 7, no. 3, 2000, pp. 969–983) as benchmarks. We report that the GM approach converges more rapidly in capturing the nonlinear dynamics of the CBC than the continuum GENE code (Jenko et al., Phys. Plasmas, vol. 7, no. 5, 2000, pp. 1904–1910) when comparing the number of points representing the velocity space. Increasing the velocity dissipation improves the convergence properties of the GM approach, albeit yielding a slightly larger saturated heat flux. By varying the temperature equilibrium gradient, we show that the GM approach successfully reproduces the Dimits shift (Dimits et al., Phys. Plasmas, vol. 7, no. 3, 2000, pp. 969–983) and effectively captures its width, which is in contrast to the gyrofluid framework. In the collisional regime, the convergence properties of the GM approach improve and a good agreement with previous global particle-in-cell results on transport is obtained (Lin et al., Phys. Rev. Lett., vol. 83, no. 18, 1999, pp. 3645–3648). Finally, we report that the choice of collision model has a minimal impact both on the ion temperature gradient growth rate and on the nonlinear saturated heat flux, at tokamak-relevant collisionality.
Publisher
Cambridge University Press (CUP)