pems of auxiliary equipment included explosive aquibe, electronic timing circuitry, a parachute system, a closure system for the sampling section, a radio beacon, and a dye marker. gamed plastic inserts were fitted !~ 4 the nose wections to provide additional buoyancy. The explosive squiba were used to remove the conical nose tip, thereby opening the sampling orifice, and to jettison the propulsion unit. The electronic timing circuitry initiated the opening of the orifice, disconnected the propulsion unit, ejected the parachute, closed the sampling section and activated the radio beacon. The parachute system consisted of a pilot chute, a pilot chute shroud cutter, and the main canopy. The pilot chute was withdrawn from its compartment when the propuiaion section was jettisoned but remained attached by shrouds to the nose section until the latter had slowed down to a speed that would not cause damage to the main canopy. At this time, the pilot chute shrouds were cut free from the nose cone, and the main canopy was withdrawn from the nose section by the pilot chute shrouds, which were still attached to a bag containing the large parachute. The front closure of the sampling unit, made by a ball joint, and the aft closure, consisting of a cone and O-ring seal, were closed after sampling. The radio beacon was activated at launch time so that search craft equipped with radio direction finders could locate the nose sections. Figure 2.3 is a view of a battery of six rockets assembled for firing. 2.2.2 Aircraftborne Sampiers. Three different types of equipment were utilized to obtain the samples discussed in Sections 1.3.3 and 1.3.4. Units of the kind illustrated in Figure 2.4 were used for collection of the cloud particle samples needed for the radiochemical work. These samplers were stainless steel shells of parabolic shape fitted with intake butterfly valves, which were open only during the sampling runs. They were installed at the forward end of both the right and left wing fuel tanks of the B-57D’s. The particles were collected on a 24-inch- diameter filter paper, which was supported by a retaining screen located near the aft end of the unit. The coincident sampler was designed so that both the gas and particle samples would be taken from the same volume of the cloud. Air was drawn through a desiccant section and a filter section by a circulating pump and then forced under pressure into a sample bottle. Fig- ure 2.5 shows the intake and desiccant-filter sections, and Figure 2.6 is a photographof the compressor pumps and gas bottles. These samplers were mounted on both sides of the B-5S7D fuselage toward the rear of the aircraft. The WB-50’s used for the fallout sampling were equipped with Air Force Office of Atomic Energy (AFOAT-1) standard E-1 filter assembly. Figure 2.7 ig a view of a WB~-50 with the filter foil installed on top, nearly over the rear scanner’s position. Figure 2.8 shows the filter screen removed from the foil with a filter paper in one side. The foil was sealed by sliding doors in front and back of the filter screen except during the sampling periods. 2.2.3 -Possible Errors in Sampling. Polydisperse aerosols contain an aggregate of particles whose sizes are arranged in accordance with a characteristic frequency distribution. When the aerosol is sampled under ideal conditions, the ratios of the numbers of particles in the various size ranges will be preserved unchanged in the collector. However, a departure from the initial size distribution may be encounteredif the collecting device has a dimensional bias (non-isoki- netic condition) or if some of the particles are broken up during the sampling operation. Isokinetic samplingconditions will be achieved with a filtering device moving through the aerosol at subsonic speeds, if the air velocity into the intake of the filter is identical with the flow rate past the outside. As used in Project 2.8, both the wing tank and coincident samplers were close to isokinetic, because the velocity ratios were respectively 0.8 (or greater) and 0.7 to 0.9. However, in a few cases, the calculated velocity ratios for the coincident units were much less because of: malfunction of the sampling equipment (Appendix B). The E-1 sampler used on the WB-50’s was poor isokinetically, but this was considered to be immaterial for height line sampling where the particles in a given region should be fairly uniform in size. Samplers, such as the project rockets, which move at supersonic speed aerosol, are expected from aerodynamic theory to be unbiased. 23 with respect to the