Understanding the Sensitivity of Detachment location to External Control on JET

In this thesis, comparisons of predictions of detachment front location sensitivity model (‘DLS model') were made in one L-mode (no N2-seeding), and three H-mode detached discharges (N2-seeding). Although the DLS model and experiment are qualitatively in agreement the model overestimates the sensitivity of the movement along a field line for a given amount of change in two control variables – upstream density, nu and the power entering the scrape-off layer, PSOL. The quantitative mismatch between the DLS model and experiment is not unexpected given the simplicity of the DLS model. Under the assumption that the exponents of nu & PSOL in the DLS model could be incorrect, new exponents were derived for both control variables from the L-mode experimental data in periods of the L-mode case where first only nu, and then both nu & PSOL were varied. The new `empirical' exponents were reduced from the DLS predictions by a factor of approximately 2.5. Those `empirical' exponents in the DLS model were then applied to the prediction of divertor impurity concentration, fz for two H-mode cases which had divertor N2-seeding show reasonable consistency with the variation in core Zeff and variations in PSOL.

As part of this work, a measure for the detachment front location was developed based on the transition of the ratio of Dgamma/Dalpha emissivities from high (recombination-dominated) to low (excitation-dominated). The detachment extent along a field line, $\hat{z}$, was found to be inversely proportional to the outer divertor total ion current which may lead to new insights to the relationship between detachment and target ion current loss. The quality of this relationship was unexpected due to many nonlinear divertor processes that could have affected it.