Density behavior during electron internal transport barriers in TCV fully non inductive discharges

Experimental results of particle transport during electron internal transport barriers (eITBs) achieved in the TCV device (Tokamak à Configuration Variable), in fully non inductive discharges, are shown. A database of steady state density profiles is constructed and studied in detail. An interesting relation between density and temperature scale lengths is observed. In the fully non inductive eITB scenario, the density profile is observed to develop a localized gradient, correlated via an almost constant factor with the electron temperature scale length up to the barrier foot. Sensitiveness of this behavior to the degree of the local confinement improvement is demonstrated by the study of the density response in eITBs with applied inductive current perturbations. The results shown have an impact also on the predictions about impurity transport in eITBs, and generally for fully non inductive operations in future machines. Introduction Particle transport observed in Ohmic L-mode and with central electron cyclotron heating has been investigated in several works ([1],[2] and references therein). Understanding the behavior of density profiles in eITB scenarios with fully non inductive current density, is an important issue both experimentally and theoretically. In TCV, thanks to the powerful and versatile electron cyclotron heating system (up to 2.7 MW at 82.7 GHz in X2 mode) high performance eITBs are routinely produced [3]. The fully non inductive discharges presented here are characterized by the total replacement of the Ohmic current with off-axis electron cyclotron current drive (ECCD) and a high fraction of bootstrap current produced thanks to the improvement in the global energy confinement. The resultant current density profile is hollow and peaks off-axis. Experimental observations: overview The database under consideration covers the following parameter range: Ip ∼ 70− 100kA, PEC = 0.9−2.3MW , ρEC ∼ 0.3−0.7, q95 ∼ 8−17 and < ne >∼ 0.2−1.1[1019m−3]; in these plasmas collisionality is low, i.e. νe f f < 0.1, where νe f f ≡ νei/(εcs/a), cs is the ion sound velocity and a is the plasma minor radius. If not explicitly shown, estimated error bars on the gradients are about 20%, while for profiles they are 5− 10%. Radial dependent quantities are averaged in 0.2 < ρ < 0.6 where ρ is the normalized poloidal flux coordinate. Fig. (1a) shows that the normalized electron temperature and density gradients are proportional with a ratio approaching ∼ 0.5 as the eITB becomes stronger; the strength of the eITB is defined as simultaneous high R/LTe ≡ R0 <|∇Te|> Te and high HRLW ≡ τ E τRLW E , where τEXP E is the energy confinement time and it is compared with the Rebut-Lallia-Watkins scaling [4]. In fig. (1b) we compare the ratio of the logarithmic gradients, 1/ηe ≡ LTe/Lne , with the Ohmic phase. 9 10 11 12 13 14 15 16 4 5 6 7 8 9 R/L T e R /L n e H RLW < 3 3 < H RLW < 3.5 H RLW > 3.5 4 6 8 10 12 14 16 0 0.2 0.4 0.6 0.8 1 R/L T e 1 /η e Ohmic H RLW < 3 3 < H RLW < 3.5 H RLW > 3.5 Figure 1: a) R Lne vs R LTe for different values of HRLW , the magenta line has a slope of 0.5; b) 1/ηe ≡ ∂ lnne ∂ lnTe vs R LTe for different HRLW , Ohmic mode values span a much wider range (magenta circles). 0 0.2 0.4 0.6 0.8 1 1 1.5 2 2.5 3 ρψ n e / n e ( 0. 85 ) averaged on t = [1.3 1.7]s