Signal-to-noise ratio improvements in laser flow diagnostics using time-resolved image averaging and high dynamic range imaging

Abstract

Two alternative image readout approaches are demonstrated to improve the signal-to-noise ratio (SNR) in temporally resolved laser-based imaging experiments of turbulent phenomena. The first method exploits the temporal decay characteristics of the phosphor screens of image intensifiers when coupled to an interline-transfer CCD camera operated in double-frame mode. Specifically, the light emitted by the phosphor screen, which has a finite decay constant, is equally distributed and recorded over the two sequential frames of the detector so that an averaged image can be reconstructed. The characterization of both detector and image intensifier showed that the technique preserves the correct quantitative information, and its applicability to reactive flows was verified using planar Rayleigh scattering and tested with the acquisition of images of both steady and turbulent partially premixed methane/air flames. The comparison between conventional Rayleigh results and the averaged ones showed that the SNR of the averaged image is higher than the conventional one; with the setup used in this work, the gain in SNR was seen to approach 30 %, for both the steady and turbulent cases. The second technique uses the two-frame readout of an interline-transfer CCD to increase the image SNR based on high dynamic range imaging, and it was tested in an unsteady non-reactive flow of Freon-12 injected in air. The result showed a 15 % increase in the SNR of the low-pixel-count regions of an image, when compared to the pixels of a conventionally averaged one.