One step was the optimization of the optic system. A reduction, e.g. of the beam diameter in the measuring volume first of all decreases the measuring rate. On the other hand, the light intensity in the measuring volume is consequently increased and thus the threshhold at which the smallest particles in the measuring volume may be detected is lowered. As the number of small particles in the usual particle distribution is very much greater than that of the big particles, a reduction in the beam diameter results in a higher measuring rate and thus a shortening of the measuring time. A reduction in the beam diameter was possible by optimizing the beam path and using special objectives. The measuring time was consequently reduced to about 1/2 or 1/3. Furthermore this also resulted in the spatial resolution of the L2F system being improved so that measurements up to a minimum distance of circa 0.2 mm normal to the wall are possible today.
A new electronic concept with automized measuring procedure has proved to be particularly advantageous in application in the rotating turbomachinery components. By increasing the number of measuring channels it was possible to simultaneously measure 16 different segment positions instead of taking measured values from one single position as had been the case up to then. A further approach to a reduction in the measuring time resulted from the following idea: If in the evaluation of the measured data we limit ourselves to the magnitude and the direction of the mean flow velocity and the degree of turbulence - that is to say disregard the determination of the Reynold's shear stress and high order moments of fluctuation velocity - it can be shown that at almost constant measuring accuracy the number of individual measurements required per measuring position may be reduced considerably. The two measures together, namely reduced data uptake in connection with the new automated electronics resulted in a further reduction in the measuring time to 1/4 to 1/5 of the value obtained after the first step.
A further reduction in the measuring time of about the same order of magnitude was obtained in a development step recently carried out. In the L2F technique it is found that measuring time increases with increasing flow turbulence as a result of the increasing angle area, due to increasing turbulence , in which the flow vector fluctuates. Thus on the one hand the number of angle positions in which the beam plane has to be adjusted in a measuring procedure increases, on the other hand the frequency of successful individual measurements decreases. The reason for this is the specified angle range determined by beam diameter and beam separation and within which valid measuring values may be registered. Should the beam separation in the measuring volume be decreased then this increases the angle range at constant beam diameter. Consequently the measuring frequency is increased and additionally the number of necessary angle positions of the beam plane is decreased, i.e. the measuring time decreases. However the measuring error increases with the beam separation becoming smaller and smaller.It can now be shown that when a maximum measuring error of e.g. 1% is assumed, the measuring time can be minimized when the separation of the beams is adjusted to the flow turbulence. Thus with a constant beam diameter of 10 mm, e.g. at 1% turbulence the best separation is 350 mm and at 10% turbulence 70 mm. In both cases measuring times are the same. If a measurement was carried out with a system with a beam separation of 350 mm in a flow with 10% turbulence then the measuring time would almost be fivefold. Thus the minimization of the measuring time requires the adjustment of the beam separation to the degree of flow turbulence.
A concept for an optical device incorporating continuous beam separation variations is hardly possible. A gradual variation in the beam separation could be achieved with exchangeable beam dividers (Rochon prisma), but this is coupled to expensive constructions. The solution found also enables stepwise beam separations and by using fiber optics results in a simply assembled, very stable optical head of small dimensions.
The light of an Argon laser operating in multi-color mode is coupled in the optical fiber LI and conducted to the optical head (Figure 2). The diverging multi-color laser light emerging from the fiber is collected by lens L1, aligned in parallel, and arrives at a beam divider prism, in this case a dispersion prism. Here the various colors of the Argon laser experience different angular deflections, so that with the aid of lens LS1, five parallel, varicolored beams with differing separations - see detailed graph - are projected in the measuring volume. The multi-colored scattering light emitted by the particles traversing these beams is collected by the outer area of lens LS1 and sent through the same dispersion prism where the various colors are again deflected such that the light produced at various places in the measuring volume is projected in one single point with lens LS2 and thus may be coupled into the optical fiber LII which functions as a spatial aperture. This fiber conducts the scattering light to a color separation and detector unit. Here a dispersion prism separates the colors into the five colors which are then each conducted in an assigned fiber to a switching arrangement. This allows the freely selectable assignment of the various colors going to the photodetectors which provide the start or stop signals for the time-of-flight measurements. In this way, by selection of a certain color combination the associated beams in the measuring volume are used for the measurement and consequently the associated beam separation is determined. The optical system is designed such that beam separations between 70 mm and 400 mm may be selected. A special achromatic design of the optic system is required to minimize the differences in the axial position of the varicolored focal points. The successful testing of the device confirmed the forecast reduction in the measuring time at high degrees of turbulence. As a result of all the measures described in this section the measuring time has been reduced on the whole to 1/50. Today, for a single measurement point e.g. in wind tunnel investigations, it requires a mere 15 seconds.