Experimental results on impurity transport in tokamaks are based on various techniques. We study here how the choice of the injection technique and of the analysis method influences the results. We have used three different injection techniques available in Tore Supra: laser blow-off, gas puff and supersonic molecular beam injection. We show that the long time duration of the gas puff injection compared with particle confinement time provides very limited information. The laser blow-off technique and supersonic pulsed injections give satisfactory results for diffusion but low quality convection estimates, presumably because the fast source term quenches the role of convection in the continuity equation. The best method is shown to be the combined analysis of supersonic pulsed injections and continuous puffing of a gaseous species. We obtain convection velocity profiles to an uncertainty of about 0.5 m s−1. This method is applied to ohmic, weakly sawtoothing plasmas. The diffusion coefficient is independent of the impurity charge and the convection velocity is inward. Neoclassical calculations show that these plasmas are dominated by anomalous transport. Quasilinear gyrokinetic simulations are in qualitative agreement with the above experimental results. We deduce from the simulations that convection is dominated by the curvature term, which means that no charge dependence should be expected in this situation.
Particle transport in diverted TdeV