Application of Thomson scattering at 1.06mm as a diagnostic for spatial profile measurements of electron temperature and density on the TCV tokamak

The variable configuration tokamak, TCV, in operation at CRPP since the end of 1991, is a particularly challenging machine with regard to the experimental systems that must provide essential information regarding properties of confined plasmas with strongly shaped, non-circular cross-sections. Although TCV is unique in its capacity for the study of magnetic equilibria not previously examined in modem, large tokamaks, this flexibility poses serious problems for the experimentalist who may be required, for example, to make measurements in completely different configurations from one discharge to the next. Highly shaped plasmas also render more complex, or even impossible, the application of inversion techniques for the recovery of plasma profiles based on chordal measurements which necessarily yield line averaged quantities. The importance of the energy confinement issue in a machine designed specifically for the investigation of the effect of plasma shape on confinement and stability is self-evident, as is the necessity for a diagnostic capable of providing the profiles of electron temperature and density required for evaluation of this confinement. For TCV, a comprehensive Thomson Scattering (TS) diagnostic was the natural choice, specifically owing to the resulting spatially localized and time resolved measurement. The details of the system installed on TCV, together with the results obtained from the diagnostic comprise the subject matter of this thesis. A first version of the diagnostic was equipped with only ten observation volumes. In this case, adequate spatial resolution can only be maintained if measurements are limited to plasmas located in the upper half of the highly elongated TCV vacuum vessel. The system has recently been upgraded through the addition of a further fifteen observation volumes, together with major technical improvements in the scattered light detection system. This new version now permits TS observations in all TCV plasma configurations, including equilibria produced in the lower and upper halves of the vacuum vessel and the highly elongated plasmas now routinely created (κ=2.47 is the maximum elongation achieved at the time of writing). Whilst a description of the new detection system along with some results obtained using the extended set of observation volumes are included, this thesis reports principally on the hardware details of and the interpretation of data from the original, ten observation volume system. The complexity of the TCV Thomson Scattering system can only be effectively conveyed through considerable descriptive effort and such details can be found in the earlier chapters of this work. Effort is also required if the set of discrete data points constituting the profile is to be effectively fitted over the wide range of profile shapes encountered in TCV. For this purpose, a number of analysis routines have been developed during the course of this research with which TS profile data can be reliably fitted with a minimum of user intervention. These routines are based on cubic spline interpolation within a normalized poloidal flux coordinate system facilitating the comparison of TS data with the results of other TCV diagnostics. However complex a given tokamak diagnostic may be, its primary purpose is, of course, to provide relevant data for use in understanding the results obtained from any particular experimental campaign. The hardware descriptions and data analysis techniques of the earlier chapters thus give way, in the second half of this thesis, to a series of studies dedicated to the use of TS data for physics understanding. The absence of an additional heating system on TCV throughout the duration of this research, necessarily limits the scope of such studies to the case of ohmic plasmas only. Some effort is devoted to an investigation of the extent to which the phenomena of profile consistency in ohmically heated discharges is observed in TCV. In general, if a form for the edge safety factor appropriate to shaped plasmas is adopted, the effect does appear to prevail, at least for elongations up to κ= 1.9 and for plasma triangularities in the range -0.4