Accuracy study of a 2-component Point Doppler Velocimeter (PDV)

Abstract

A two-component Point Doppler Velocimeter (PDV) which has recently been developed is described, and a series of velocity measurements which have been obtained to quantify the accuracy of the PDV system are summarized. This PDV system uses molecular iodine vapor cells as frequency discriminating filters to determine the Doppler shift of laser light which is scattered off of seed particles in a flow. The majority of results which have been obtained to date are for the mean velocity of a rotating wheel, although preliminary, data are described for fully-developed turbulent pipe flow.Accuracyof thepresent wheel velocity datais approximately + i % offullscale, whileImcarity of a single channel ison theorder of +-0.5% (ie, + 0.6 m/see and ± 0.3 m/see, out of 57 m/see, respec_ely).The observed linearity of these results is on the order of the accuracy to which the speed of the rotating wheel has been set for individual data readings.The absolute accuracy of the rotating wheel data is shown to be consistent with the level of repeatability of the cell calibrations.The preliminar/turbtdem pipe flow data chow consistent turbtdence intensity values, and mean axial velocity profiles generally agree with pitot probe data. However, them is at present an offset error in the radial velocity which is on the order of 5-i0 % of the mean axial velocity.

Results for a Two-Component Doppler Global Velocimeter

“…An offset on the order of 5 m/s in the DGV measurements has been subtracted from both cuts. The cause of this zero velocity offset is believed to be random, uncorrelated drift in the side arm temperatures of the reference and measurement iodine cells; similar zero offsets were observed in the PDV results of Kuhlman et al 3 The rms deviation of the tangential velocity in Fig. 6 from the known solid body wheel rotation varies from 1.2 to 1.8 m/s for X and Y cuts, with an average rms error of 1.5 m/s, for a series of four runs.…”

“…However, large errors in mean and rms velocities were found near the edges of the jet, probably due to uneven seeding and low signal-to-noise ratio. Similar PDV studies have been conducted by Morrison et al, 14 Roehle and Schodl, 15 and Kuhlman et al 3 More recently, McKenzie, 1 , 16 Smith and Northam, 2 Smith, 17 Irani, 18 and Irani and Miller, 19 Elliott et al, 20 , 21 Clancy et al, 22 Mosedale et al, 23 Reinath, 24 Beutner and Baust, 25 Beutner et al, 26 and Ainsworth and Thorpe 27 have also developed DGV systems. Meyers 28 describes recent improvements to his DGV systems.…”

“…McKenzie's more recent planar DGV results of the velocity of the same rotating wheel displayed a lower level of accuracy ( §2-5 m/s). 1 Two-component PDV results by Kuhlman et al 3 for a rotating wheel displayed an accuracy on the order of §0.5-1 m/s. Thus, it is clear that, in a very short time, DGV has demonstrated capability for making nonintrusive mean ow velocity vector measurements in a plane, in a variety of complex ow elds of practical signi cance.…”

“…This reduced accuracy relative to conventional laser velocimetry (LV) is primarily due to camera noise and pixel registration errors, as well as laser speckle noise for systems using pulsed Nd:YAG lasers. 1 , 2 A two-component point Doppler velocimeter (PDV) system has been previously developed by Kuhlman et al 3 A two-component, planar imaging DGV system using CCD cameras has also recently been developed,as initially describedby Naylor and Kuhlman. 4 The accuracy limits of both systems are being systematically explored, through a series of measurementsin relativelysimple ows.…”

Development and Testing of the Virginia Tech Doppler Global Velocimeter

“…The point technique relies on photo-diodes to measure light intensity through the same type of molecular iodine cells used by Meyers at NASA Langley. Later work added a second photo-diode to measure two velocity component data on a rotating wheel and achieved uncertainties less than 1 ™ [25]. Expanding on the work for the point system, Naylor developed a two component camera based DGV [20,27].…”