To determine three-dimensional velocity vectors with a laser measuring technique, a method is usually selected in which two two-dimensional systems observe the same measuring volume from different directions. The desired velocity vector is determined from a geometric transformation of the two measurement results . To obtain sufficient accuracy for all components of the velocity, the angle difference between the two systems must be at least 30°. This method requires relatively big measuring windows.
Yet access to the measurement location is limited in many cases, e.g. in turbomachinery, where the measurement location is usually only accessible via relatively small windows in the housing of the machinery. In these cases only those laser velocimeter which can operate with an extremely small solid angle, determined by the measuring window size, are applicable for the 3D velocity measurements.
Two methods on the basis of the L2F technique have been put forward to achieve this objective. In the first technique two L2F systems are placed together in one casing. As shown in the diagrammatic sketch in the upper right figure, the two systems are each set at an angle of circa 6° to the optical axis of the receiving part. Both velocimeters operate at different wavelengths so as to enable the signal received from the measuring volume to be assigned to the appropriate device. 1g stands for the green start beam and 2g stands for the green stop beam. Corresponding terms apply to the system with the blue wavelength. At point A, beams 1g and 1b intersect, at point B, beams 2g and 2b. Thus, as a result of the two systems, measuring planes located obliquely to each other are spread over the space. For flow vectors perpendicular to the optical axis the two systems each behave like a standard L2F device. However, when the velocity vector also has a component in the direction of the optical axis, then the two systems still measure the same velocity magnitude but different angles. Using the measured angle difference and the inclination of the beam axis to the optical axis, the actual 2D flow angle as well as the angle of the flow vector with respect to the optical axis can then be ascertained, resulting in complete determination of the velocity vector.
The practical implementation of this concept, which goes back to the beginning of the 80s, foundered first of all on the considerable light scattering in the image rotating prism, so that application of the device in turbomachinery was practically out of the question. But by using glass fibers it was possible to design an improved device. The disturbing influence of the image rotating prism was eliminated by an optical head which could be rotated as a whole.
When access to the measuring point is even more difficult, then the second 3D-L2F technique may be applied. The figure shows a schematic longitudinal section of the measuring volume of two L2F systems which differ from a standard measuring volume in the axial displacement of the start- and stop focus. The focussing areas usable in measurements, resulting from the light intensity distribution and the apertures in the receiving beam path, are section lined in the figure. For an assumed flow vector the resulting effectively usable lengths of the measuring volume LA, LB are indicated by dotted lines. In the case shown here, system A had a considerably higher rate of successful start - stop occurences than did system B. The normalized difference in the two observed data rates stands for not too great flow angles ( |b| < gA,B) in near linear relation to the flow angle b. The slope of the resulting calibration curve must be determined in this technique because of the unknown absolute length of the focussing area and the particle size distribution which is not previously known. To produce the measuring volume with axial displacement a similar test set-up is used as in the L2F technique with variable separations. By exchanging the front lens for a lens not corrected for the wavelengths of the laser light, the individual color beams are focussed at variable distances and thus a system is set up corresponding to for example System A. By turning the optical head 180° and exchanging the start and stop beams System B is automatically obtained. The required difference in the data rates is determined from two consecutive measurements with both arrangements and then the flow angle relative to the beam axis can be calculated from the calibration carried out beforehand. In both arrangements the same flow angle perpendicular to the flow axis, the same velocity magnitude as well as the same fluctuation values are measured.