Modelling of ECH modulation experiments in ASDEX Upgrade with an empiricalcritical temperature gradient length transport model

The relevance of the assumption of critical gradient length in electron hea t transport is tested against the results of electron cyclotron heating (EC H) modulation experiments carried out on ASDEX Upgrade, using time-dependen t transport simulations. A general chi (e) model based on a delT(e)/T-e thr eshold is proposed, which can provide two different heat pulse (HP) propaga tion regimes, with respect to the power balance (PB) diffusion coefficient chi (PB)(e) : chi (HP)(e) = chi (PB)(e) occurs if delT(e)/T-e < (delT(e)/T- e)(c), while chi (Hp)(e) > chi (PB)(e) if delT(e)/T-e greater than or equal to (delT(e)/T-e)(c). By a careful choice of the model parameters, the expe rimental T-e response to ECH modulations is simulated with excellent accura cy, and the transition between the two propagation regimes is, without ambi guity, ascribed to a delT(e)/T-e threshold. Our findings are consistent wit h theoretical expectations, as well as with experimental observations made on TORE SUPRA. This global consistency shows the relevance of the main phys ical factors involved in our model, in particular the assumption of critica l temperature gradient length.