TM No. 377 



ratios are plotted in figure 11-24. The horizontal dashed lines represent 

 the tangents of the various angles shown on the right hand side of the 

 figure j the values are read off the ordinate axis. The points clustered 

 about each tangent line are actual data points taken from a series of 

 acceleration tests. Since the ordinate is logarithmic, the spread of 

 data points, with the exception of 9 = 80 , is relatively small. Note, in 

 particular, that there is no apparent variation of the ratio with tow 

 speed. 



Thus, the ratio u/w is essentially independent of the absolute speed of 

 the velocity vector in the plane of the u and w meter. By determining the 

 value of 9 from the ratio and applying a correction to each of the velocity 

 components, one can obtain an estimate of the error of u as a function of 

 9 and of w as a function of 90° - 9 (and vice-versa). The method of 

 correction of the OMDUM III data output is discussed in the section on 

 data processing in chapter III. 



Estimation of Response Time. The response time of the OMDUM III system 

 was estimated using the 1.5-meter-diameter wind tunnel in the NUWS hydro- 

 dynamic laboratory. The system was mounted with the u meter axis centered 

 in the tunnel and parallel to the air flow. The tunnel fan was driven to 

 provide a steady flow at about 200 cm sec" 1 . A square of cardboard was 

 held flush to the upwind u meter cylinder, blocking out all air flow. The 

 cardboard was removed smartly, and the impeller output registered on a 

 strip chart recorder. 



Figure 11-26 is a plot of the time between every sixth voltage pulse 

 (which is equivalent to the period of revolution for the six bladed impeller) 

 as a function of time after the cardboard was removed. The period calculated 

 occurs at the mid-point between each sixth pulse. For the first few pulses 

 the curve is somewhat distorted because the period is changing exponentially 

 with time. However, at the outer end of the curve, the rapidity of the 

 points gives a more realistic value of the period of rotation as a function 

 of time. 



The curve in figure 11-26 is easily shown to be exponential. It can be 

 represented by the equation: 



_ftt)=^(oo)][|- e-5-*] ; (rwJ3) 



where the frequency function is 



f -. T*' (11-24) 



and f (oo) is the frequency attained by the impeller when it has come to 

 equilibrium with the steady air flow. 



Using the, curve in figure 11-26, f ( OO ) <«. 9.62 cps, and the time consant 

 in air T. is about 3.5 seconds. This time constant can be measured directly 



39 



