The first transport code simulations using the trapped gyro-Landau-fluid model

Abstract

The first transport code simulations using the newly developed trapped gyro-Landau-fluid (TGLF) theory-based transport model are presented. TGLF has comprehensive physics to approximate the turbulent transport due to drift-ballooning modes in tokamaks. The TGLF model is a next generation gyro-Landau-fluid model that improves the accuracy of the trapped particle response and the finite Larmor radius effects compared to its predecessor, GLF23. The model solves for the linear eigenmodes of trapped ion and electron modes, ion and electron temperature gradient modes, and electromagnetic kinetic ballooning modes in either shifted circle or shaped geometry. A database of over 400 nonlinear gyrokinetic simulations using the GYRO code has been created. A subset of 83 simulations with shaped geometry has been used to find a model for the saturation levels. Using a simple quasilinear (QL) saturation rule, remarkable agreement with the energy and particle fluxes from a wide variety of GYRO simulations is found for both shaped or circular geometry and also for low aspect ratio. Using this new QL saturation rule along with a new E×B shear quench rule for shaped geometry, the density and temperature profiles have been predicted in over 500 transport code runs and the results compared against experimental data from 96 tokamak discharges. Compared to GLF23, the TGLF model demonstrates better agreement between the predicted and experimental temperature profiles. Surprisingly, TGLF predicts that the high-k modes are found to play an important role in the central core region of low and high confinement plasmas lacking transport barriers.

Characterizing turbulent transport in ASDEX Upgrade L-mode plasmas via nonlinear gyrokinetic simulations

“…This underprediction ('shortfall') was observed in quasilinear gyrofluid predictions using the TGLF model, 10,11 but as it turned out, also the gyrokinetic turbulence codes GYRO 12 and GEM 13 (run at relatively low resolution) failed to recover the experimentally inferred ion heat transport level in a selected case, underpredicting it by a factor of ∼ 7 (see Ref. 2).…”

Quasilinear transport modelling at low magnetic shear

“…27 and 28) and TGLF (Ref. 29) for external toroidal velocity. Indeed, the non-linear k y j max was also recovered by calculating maxðc eff =k 2 y Þ, where c eff is a quenched growth rate taken as c eff ¼ maxð0; c k À 0:3x ExB Þ. x ExB is the effective zonal flow ExB shear rate taken from the GENE non-linear simulations, including the effect of the autocorrelation time on the effective shearing time.…”

Direct multiscale coupling of a transport code to gyrokinetic turbulence codes

“…1,10,11,12,13,14 These reduced models have provided a basic qualitative understanding of the multiscale interaction and are capable in some cases of giving good quantitative agreement with first-principles, nonlinear gyrokinetic calculations. 15,16 However, such models do not permit detailed validation studies because they do not produce fluctuation spectra or other data that can be experimentally challenged. Without careful validation, even first-principles reduced models (with no adjustable parameters) cannot be fully trusted for predictions of performance in new operational regimes, and reduced models with fit parameters adjusted for current experiments have even less credibility.…”