It must be emphasized that this is a preliminary study. The assumptions on which it is based are very idealized. For instance, the treatment of the incident angular distribution as a pillbox and the requirement that the secondary and target be sized so as to achieve 100% intercept are somewhat extreme. In practice, the optical errors are more likely to be better represented by some form of Gaussian distribution and the trade-off between concentration and intercept may yield relationships that differ in some details from those found here. However, based on the experience of applying similar arguments to the optimization of one – and two-stage parabolic dish systems (O’Gallagher and Winston, 1985, 1988b)4, we are confident that the general conclusions reached here in this study regarding achievable concentration in particular configurations will serve as an excellent design guide for the actual behavior. With this in mind, we wish to offer the following recommendations:
1. The best designs for achieving high concentrations will be those with a circular surround field and a fairly tall tower.
2. The next best designs are those with g = 6c and as small a value of 0c and corresponding tower height as are acceptable.
However, it should be noted that in the latter design, the achievable concentration would be reduced by a factor of approximately one-third.