ITER Test Blanket Module Error Field Simulation Experiments at DIII-D

Full text: Comprehensive DIII-D experiments investigated the effects of ferromagnetic error fields similar to those expected from proposed ITER Test Blanket Modules (TBMs). Studied effects included: plasma confinement; L-H transition; edge localized mode (ELM) suppression by resonant magnetic perturbations (RMPs); locked modes; energetic particle losses; and more. DIII-D experiments used a 3-coil mockup of 2 magnetized ITER TBMs in an ITER equatorial port. The largest and most prevalent effect was slowing of plasma toroidal rotation ν across the entire radial profile by as much as δν/ν0 ∼ 50%. The slow not ing is attributed to non-resonant braking by the TBM mockup magnetic field B. There was no sign of braking by resonant magnetic harmonics. These results are consistent with the near absence of resonant helical harmonics in the TBM magnetic spectrum. Changes in global confinement (δn/n0, δβ/β0, δH98/H980) were ∼ 3 times smaller than δν/ν0. The results did not depend strongly on whether b was imposed by mockup coil current variations or mockup-to-plasma distance changes. TBM effects increased with increasing βN and were smaller in L-mode. H-mode power threshold rose ≤15% within the uncertainty of the measurements. Mockup field increased the plasma sensitivity to mode locking by known, deliberately applied n = 1 error field proxies (n = toroidal harmonic number). The standard DIII-D locked mode tolerance at low β was recovered by the simple expedient of a new empirical compensation of the total n = 1 error. At high β the increased sensitivity to locking was due to TBM braking torque slowing the plasma. Velocity reduction may be the primary cause of the confinement reductions. Quantifying TBM effects on ELM suppression by n = 3 RMPs was difficult, because small TBM B had little effect on ELMs but larger B (∼ 3 times one ITER port's error) slowed the rotation of neoclassical tearing modes in the high-β test plasmas, which often produced a back-transition to L-mode. It may be possible to extrapolate a number of these results to an ITER with 2 TBMs in a single port, but extrapolating to 6 TBMs in 3 ports is less clear. Effects related to low-n TBM harmonics may be correctible by conventional error field correction. Work supported by USDOE DE-FC02-04ER54698. (author)