Measurements of injected impurity assimilation during massive gas injection experiments in DIII-D

Abstract

Impurities (H2, D2, He, Ne or Ar) injected into steady (non-disrupting) discharges with massive gas injection (MGI) are shown to mix into the plasma core dominantly via magnetohydrodynamic activity during the plasma thermal quench (TQ). Mixing efficiencies of injected impurities into the plasma core are measured to be of order 0.05–0.4. 0D modelling of the experiments is found to reproduce observed TQ and current quench durations reasonably well (typically within ±25% or so), although shutdown onset times are underestimated (by around 2×). Preliminary 0D modelling of ITER based on DIII-D mixing efficiencies suggests that MGI will work well in ITER with regard to disruption heat load and vessel force mitigation, but may not collisionally suppress runaway electrons.

Disruption mitigation by massive gas injection in JET

“…Figure 18 The data shown in figure 18 shows an increase of the efficiency with lower 95 , whereas the plasma current has less impact on the efficiency if is adjusted such to keep 95 constant. This dependence on 95 is consistent with what has been seen at DIII-D from direct density measurements [9]. However, it remains an open question if the flattening of the current profile shows a dependence on 95 as well, which could also partly explain the observed trend.…”

Fuelling efficiency of massive gas injection in TEXTOR: mass scaling and importance of gas flow dynamics

“…This value seems to be higher than 0.2 typically found in DIII-D [22]. However, here the efficiency was normalized to the number of atoms delivered before the thermal quench, in contrast to DIII-D, where the atoms delivered up to the middle of the current quench are considered.…”

The ITPA disruption database

“…The ability to transport impurity ions into the core is measured by the fueling efficiency, here defined as reference [45], also used in reference [13], which is readily measured by multiple devices. This is in contrast to the fueling efficiency € Y mix defined in [47], which utilizes estimates of the average impurity charge state during the CQ to calculate the impurity ion population in the plasma, and therefore directly measures the ion assimilation. However, this second technique requires complex spectroscopic techniques to resolve the ion charge state, and hence is not included within the IDDB at this time.…”

“…antennae) with shorter time constants may be affected. is likely due to the fact that the fastest IDDB current quenches occur in DIII-D, which has entirely carbon plasma facing components (PFC) that are known to emit significant amounts of carbon impurity into the plasma during the thermal quench (TQ) [47]. The distribution for natural disruptions skews significantly towards longer € Δt cq for beryllium first wall and tungsten divertor devices, as has recently been reported for the ITER-like wall in JET [51] (data not presently available in IDDB) , and is expected to do so in ITER as well.…”