Global Doppler frequency shift detection with near-resonant interferometry

Abstract

The recent development in measuring 2D Doppler shift distributions for flow velocimetry using the dispersive properties of atomic line filters is presented. On the basis of velocity field measurements on a subsonic jet flow and a tip vortex flow in a medium-sized wind tunnel, the technique was assessed. Atomic line filters near a resonant transition combine imaging capabilities with a sharp frequency cutoff and an associated region of strong anomalous dispersion. While conventional Doppler global velocimetry relies on the absorption of the filter to convert frequency shifts to intensity variations, near-resonant interferometry uses its dispersion to detect frequency shifts as phase changes in an interference pattern. In the present setup, an iodine vapor cell in an imaging Michelson interferometer is used. With the illuminating laser frequency tuned near a resonant transition, the cell’s dispersion converts the frequency content of the field of view into a distortion of the carrier-fringe pattern recorded at the image plane of the interferometer. The phase distribution in the fringe images is reconstructed by filtering the individual images with a 2D Gabor filter pair tuned to the spatial frequencies of the basic carrier-fringe pattern. The post-processing is concluded with subsequent phase-unwrapping and subtraction of the carrier reference fringe phase. The method and the setup were demonstrated and calibrated experimentally on a rotating disc. The capability of the technique to operate in a real experimental environment was validated in a free subsonic jet and a tip vortex flow behind a wing section in a medium-sized wind tunnel facility. The measurements were found to be in generally good agreement with the theoretically predicted system characteristics and the reference measurements. As with other Doppler global techniques, the stability of the pulsed laser system and the secondary scattering in the test volume were identified as the main error sources.