Collisionality and shear dependences of density peaking in JET and extrapolation to ITER

Abstract

Results from an extensive database analysis of JET density profiles in stationary conditions show that the density peaking factor ne0/⟨ne⟩ in JET H modes increases from near 1.2 at high collisionality to around 1.5 as the plasma collisionality decreases towards the values expected for ITER. This result confirms an earlier observation on AUG. The density peaking behaviour of L modes is remarkably different from that of H modes, scaling with overall plasma shear as (ne0/⟨ne⟩ ∼ 1.5li), independently of collisionality. H-mode density profiles show no shear dependence, except at the lowest collisionalities. No evidence for LTe, LTi, ρ* or β dependences has been obtained. Carbon impurity density profiles from charge exchange spectroscopy are always less peaked than electron density profiles and usually flat in H modes. The peaking of the electron density profiles, together with the flatness of the impurity density profiles, are favourable for fusion performance if they can be extrapolated to ignited conditions.

Density profile peaking in JET H-mode plasmas: experiments versus linear gyrokinetic predictions

“…That gives us ν eff (ITER) = 0.19 and n 0.2 / n e (ITER) = 1.42-1.55, R/L n (ITER) = 2.0-2.8, depending which of the derived scaling expressions is used. The prediction for the peaking factor is in agreement with the extrapolations done in previous publications [2][3][4] despite being based on different datasets and using slightly different regression variables than here. The expectations for R/L n values are lower in this work (2.4 ± 0.4 instead of 4.0 ± 1.0), but as stated before, the value used in this work is calculated over the large density profile range and can be slightly biased with respect to definitions used in other publications.…”

“…That can explain a dramatic difference between the density peaking observed in JET and AUG H-mode plasmas and ECRH heated H-mode case on TCV [15] (with T i /T e 1 and vanishing core particle source), where density peaking is considerably higher for the same collisionalities, compared with the scalings derived in [2][3][4].…”

“…It has already been shown in earlier experiments that density peaking in H-mode plasma scales with the logarithm of the effective collisionality [1][2][3][4][5][6]. The same tendency is also clearly seen in the recent JET plasmas studied in this paper (see figure 2) with an additional feature: around ν eff ∼ 0.5 there appears to be an inflexion point where the collisionality dependence disappears, which is well visible on the n 0.2 / n e measurement and a little less pronounced on R/L n .…”

Collisionality dependence of the quasilinear particle flux due to microinstabilities

“…Density peaking in tokamak plasmas has been shown to decrease with increasing collisionality in ASDEX Upgrade [1] and JET (Joint European Torus) [2] H-modes [3][4][5].…”

Comment on “Turbulent equipartition theory of toroidal momentum pinch” [Phys. Plasmas 15, 055902 (2008)]

“…R / L N is defined positive for the usual peaked density profiles ͑R / L N ϵ −R ٌ n / n͒, and the density gradient in the TEP theory, therefore, leads to an outward flow of parallel momentum when the density profile is peaked. TEP alone can hardly be expected to lead to peaked rotation profiles for L-mode or low collisionality H-mode plasmas because the density gradient is observed 6,7 and predicted 8 to be R / L N Ϸ 3 -4 under these conditions, such that RV / Ϸ 0. Angular momentum profiles are peaked, but the peaking is mainly due to the peaked density whereas the rotation profile would be flat.…”