862 BELL SYSTEM TECHNICAL JOURNAL 



In Fig. Qa. one is assumed to be looking into the reflector along the sym- 

 metric axis. The efi'ective area is represented by the shaded hexagon. 

 Evidently, the effective area of the modified reflector is identical to the 

 effective area of the original triangular reflector ABC. Therefore, for this 

 particular aspect, the effective area has not been changed by the addition 

 of the material at the corners. 



Figure 9b is a view of the reflector at = 30°, = 0°. Again, the shaded 

 region represents the effective area, and the parallelogram abed is the effec- 

 tive area of the reflector before modification. The modification has evi- 

 dently introduced a substantial gain in effective area for this aspect. A 

 graphical com.parison of the effective areas of Figs. 9a and 9b shows them 

 to be of comparable magnitude. 



With the dimensions defined as in Fig. 10, a corner reflector was con- 

 structed with a= b= 17". The response curves of this reflector are plotted 

 in Fig. 11, along with the curves of the ordinary triangular reflector. A 

 substantial improvement in response is exhibited by the com^pensated re- 

 flector. In the region extending out to 30° from the axis, the response level 

 varies by no more than a couple of decibels. The response appears to rise 

 slightly in the vicinity of 20°. This could, perhaps, be reduced by a more 

 appropriate shaping of the sides of the reflector. 



The variation of the response curve with the ratio ^ has been studied 

 briefly. It appears that a value of ^ = 1 is about right, for the 30° contour 

 to equal the axial response. If - < 1, the reflector will only be partially 

 corrected; if ^ > 1, it will be overly corrected. In the uncorrected reflector 

 with triangular aperture, a = 0- 



If b/a = CO , that is, if a = 0, one would expect to obtain a response curve 

 having a minimum value on the axis and rising to a maximum on either side. 

 A reflector having these properties is illustrated in Figs. 12a and 12b. 

 Again Fig. 12a is the axial aspect, whereas Fig. 12b is the 30° aspect. In 

 the former, the effective area should be zero; in the latter, it has the value 

 represented by the shaded portion. A reflector of this kind, in which b = 

 34" and a = 0, was constructed and tested. The experimental results are 

 shown in Fig. 13. The minimum is, perhaps, not as low in value as expected 

 because of residual reflections from the support upon which the reflector was 

 mounted. As expected, however, the curve passes from a minimum on the 

 axis to maxima on either side. 



The above examples serve to illustrate some of the results which can be 

 realized with trihedral reflectors. We have seen that the response character- 

 istic can be controlled by appropriate modifications of the geometrical con- 

 figuration of the aperture. 



Experiments were performed in order to determine the reduction in echo 

 caused by errors in the internal angles at the corner. 



