To reduce the effect of ~rying fluid heads and a corresponding
variation in valve calibration, a ballast reservoir mounting was devised
to limit the head values. This ballast reservoir, after several modifica-
tiona, consists of a spun alwninum tank wi til filter, mounted on 18-inch
legs. It is shown in Figure 15. The legs serve as supports for the other
control units and a head of at least one foot is provided by tubing to the
autODI8.tic ballast valve. The capacity of the reservoir is approximately
five gallons. Figures 16 and 17 show the complete ballast release assembly.
One other system of altitude control may be mentioned. This is the
4
method used by Korff and others to roughly approximate constant level
flights for cosmic ray investi!ations. A numberat meteorological balloons
are inflated until they will just support tl:e flight load. A ff!IW other balloons
are added to the train to give a free lift appropriate for the desired rate
of rise (see Computations, Section 3). At some time after release these
"lifter" balloons burst due to over-inflation, or are released by a pressure
or time-activated mechanism. lf the original balance was correct, and the
effects of superheat and diffusion can~el each other, the cluster of cells
may float. When one or more of the balloons breaks, or leaks excessively.
the train will descend. Although this method was used in early experimental
flights it proved to be useful only as a stop-gap method of carrying gear
aloft for test purposes. No modification of this basic technique seems likely
to produce even a consistant flight pattern due to the uncertainty of properties
and behavior of these inherently unstable balloons.
C. Altitude Detennination
In order to evaluate the performance of the basic control
apparatus, an investigation of pressure-measuring equipment and telemetering
gear has been made. The problems of measuring upper-air conditions in general
- 10 -