Multi-scale/Multi-field Turbulence Measurements to Rigorously Test Gyrokinetic Simulation Predictions on the DIII-D Tokamak

Full text: This presentation describes the overall progress in rigorously testing gyrokinetic turbulence simulation through a series of carefully designed experiments performed within the Transport Mode Validation Task Force at DIII-D. These experiments take full advantage of the diagnostic, heating, and plasma control capabilities available at DIII-D which provide significant constraints for testing and validating gyrokinetic simulations. A unique array of multi-field, multi-scale turbulence measurements has been utilized to study a range of target plasmas with excellent spatial coverage (r/a ∼ 0.55 - 0.85). New measurements include local, wavenumber-resolved TEM-scale n-tilde, turbulence flows, density-temperature fluctuation crossphase, as well as previously available ITG and ETG scale ne and low-k Te fluctuations. New results include: 1) Agreement between neTe crossphase measurements and gyrokinetic predictions when ECH heating is used to vary Te; 2) Tests of electron mode physics by making ITG modes sub-dominant to electron modes (TEM and ETG) show that intermediate TEM/ETG-scale n-bar increases with increasing a/LTe as predicted. However, higher a/LTe levels show a surprising and unpredicted opposite response; and 3) In elongation (K) scans the measured thermal transport and low-k fluctuations (ne and Te) decrease with increasing k consistent with general predictions. These comparisons, although still in the early stages, are indicating some general observations. Predictions of transport and fluctuation levels in the middle core region (0.4 < r/a < 0.75) are often in better agreement with experiment than those in either the outer region (r/a ≥ 0.75), where edge effects may become important, or with those in the inner region (r/a ≤ 0.4) where some of the scale length assumptions in the gyrokinetic prescription may become less reliable. Further, simulations of plasmas which are not dominated by low k turbulence require extension to successively higher k ranges perhaps indicating the need for a fully coupled low through high k simulation. In summary, careful comparison of experiment and simulation has improved the design of rigorous validation experiments as well as identifying the critical importance of utilizing accurate synthetic diagnostics. Work supported by USDOE DE-FG02-08ER54984, DE-AC05-76OR00033, and DE-FC02-04. (author)