Conceptual design of multi-chord interferometer for the Versatile Experiment Spherical Torus (VEST)

Abstract Interferometers based on phase measurement technique have been widely used to measure the plasma line integrated density (LID) in the magnetic fusion devices. Phase jump is an inherent drawback of phase measurement techniques. Once a phase jump occurs, phase errors should be compensated for by referring the complementary diagnostics. In a multi-chord interferometer used to measure plasma density profiles, phase errors can be compensated for using the phase relationship between adjacent receivers. To utilize the phase relationship between receivers, microwaves emitted from a single transmitter are spatially distributed to all receivers. The total phase changes are tracked by spatially accumulating the phase differences between adjacent receivers. Since temporal accumulation of the phase changes is not required, phase measurement failure due to phase jumps does not affect subsequent measurements. In multi-chord interferometers, phase loss due to beam refraction is also a serious problem because the beams are incident on the plasma at oblique angles. The beams are refracted due to the plasma density gradient. In a multi-chord interferometer with a single distributed transmitter, the phase loss problem is automatically solved because some of the microwaves always enter the receivers via curved beam paths, but interpretation of the measurements is problematic because the actual beam paths differ from the straight paths in vacuum. An algorithm is developed to find the actual curved beam paths by iterating four procedures: reconstructing density profile using tomography from the previously estimated LID, ray tracing beam path in the reconstructed density profile, estimating LID with the ray traced beam path information from phase measurements, and averaging the estimated LID with the previous LID. In this paper, the principles of single transmitter multi-chord interferometry are demonstrated by analyzing synthetically generated data and the algorithm of iterative tomography is described in detail.