Density peaking in JET-determined by fuelling or transport?

Core density profile peaking and electron particle transport have been extensively studied by performing several dimensionless collisionality (upsilon*) scans with other matched dimensionless profiles in various plasma operation scenarios on the Joint European Torus (JET). This is the first time when electron particle transport coefficients in the H-mode have been measured on JET with high resolution diagnostics, and therefore we are in a position to distinguish between the neutral beam injection (NBI) source and inward electron particle pinch in contributing to core density peaking. The NBI particle source is found to contribute typically 50%-60% to the electron density peaking in JET H-mode plasmas where T-e/T-i similar to 1 or smaller and at upsilon* = 0.1-0.5 (averaged between r/a = 0.3-0.8), and being independent of upsilon* within that range. In these H-mode plasmas, the electron particle transport coefficients, D-e and v(e), are small, thus giving rise to the large influence of NBI fueling with respect to transport effect on peaking. In L-mode plasma conditions, the role of the NBI source is small, typically 10%-20%, and the electron particle transport coefficients are large. These dimensionless upsilon* scans give the best possible data for model validation. TGLF simulations are in good agreement with the experimental results with respect to the role of NBI particle source versus inward pinch in affecting density peaking, both for the H-mode and L-mode upsilon* scans. It predicts, similarly to experimental results, that typically about half of the peaking originates from the NBI fuelling in the H-mode and 10%-20% in the L-mode. GENE simulation results also support the key role of NBI fuelling in causing a peaked density profile in JET H-mode plasma (T-e/T-i similar to 1 and upsilon* = 0.1-0.5) and, in fact, give an even higher weight on NBI fuelling than that experimentally observed or predicted by TGLF. For the non-fuelled H-mode plasma at higher T-e/T-i = 1.5 and lower beta(N) and upsilon*, both TGLF and GENE predict peaked density profiles, therefore agreeing well with experimental steady-state density peaking. Overall, the various modelling results give a fairly good confidence in using TGLF and GENE in predicting density peaking in quite a wide range of plasma conditions in JET.