Laser doppler anemometers are non-invasive, linear, and inherently precise. Calibration, in the usual sense, is not necessary; length and frequency measurements suffice to establish velocity at a spatial point. Measurements were made at points in the cross-sections of two square ducts containing water flow. The points were selected in conformance with a numerical integration scheme to be used for volumetric flow rate determination from the velocity measurements. The experiments were performed in a primary calibration facility at flows up to 1.25 m3/sec using ducts with sides 46 and 92 cm. The anemometer, operating in forward scatter differential mode with a 15 mw He Ne laser, was positioned with a special traversing frame. Windows in the ducts allowed transmission of the beams into the flow and reception of scattered light. Two grid patterns, 4 x 11 and 11 x 11, were used so that 44 and 121 velocities were measured for each test. A total of eight tests were conducted covering a Reynolds number range from 1.1 to 3.9 X 108. After accounting for errors due to the discreet integration scheme of 0.61% and 0.13% for the 4 x 11 and 11 x 11 schemes, respectively, comparison with the calibration facility indicated extreme errors of +0.81/-0.16 and +0.84/-0.61. The major limitation of the set-up used was the time required to move the anemometer and obtain a new velocity value. It was pointed out that either better mechanical positioning or optical scanning could be employed to reduce the time required for a flow determination.
Available at: http://works.bepress.com/wdurgin/38/