166 
GOETZ AND PREINING 
Fic. 1—View of the aerosol spectrometer, partly dissembled; component 
parts (left to right): 
chassis, cup, housing shell, bearing, motor-rotor as- 
sembly, foils (one showing aerosol spectrum) (Courtesy of the Instrument 
Development & Mfg. Co., Pasadena, California) 
dimensions of the helix yield the following 
F = 2.75 X 1030N 
v, = 12.9F = 3.55 X 10°ON 
To = 3.15/F 
Nr = 73F 
Vinin = Oree oe Ome 
Ymax = 1.52 X 10-1N? 
The value of F can be determined at the inlet 
by a flow meter which may not impart a flow 
resistance, while the outflow W may have to be 
restricted to yield the same F value at the inlet, 
in order to compensate for dynamic pressure 
conditions between rotor and housing [Goetz, 
1959]. 
The principle of the size classification of the 
airborne particles in the A.S. is based on the 
variation of the time 7, with the diameter d of 
the particles, required for their radial transport 
to the foil across the laminar air flow in the cen- 
trifugal field. If (Zz) is the distance from the 
channel entry along the outer helical channel sur- 
face, where no more particles of a size greater than 
d are deposited, it is obvious that (Za) must in- 
crease with decreasing d for the same values of O 
and N. Hence polydisperse aerosols with a small- 
est particle size d form a deposit along L which 
terminates for (L = Luz), provided that N and O 
are such that (La < L.). Similarly it follows that 
for the same d value (La) decreases with increas- 
ing N orzo, also that (Zz) indicates d, if the func- 
tion d(La) is known. 
Figure 3 represents the function d(Zz) for a 
variety of N and O values, selected for the uni- 
form coverage of the range of d and L, as ob- 
tained from calibration with monodisperse poly- 
styrene latex aerosols of eight different sizes: they 
represent straight lines in logarithmic coordinates, 
with a slope increasing with N and decreasing 
with O. 
The size ranges (1.2 » > d > 0.08 wp), indi- 
cated by solid lines, were calibrated with latex 
aerosols, while the hatched sections indicate the 
d ranges for which no latex preparations are avail- 
able. 
Similarly uncertain is the validity of the lear 
relation for La < 4 cm because of the initial dis- 
turbance of the flow after entry into the channel 
and at the approach to the orifices, that is, for 
La > 38 em. 
This limits the range of d for the N and O 
values tested to dnax < 3 yu for N = 6000,O0 = 
1.5, and dmin > 0.03 wu for N = 24,000; O = 0.5. 
The d range accessible to analysis embraces thus 
the factor of about 100, and the corresponding 
range of particle masses of about one million. 
Table 1 correlates the values of F, v,, 70, 
Ymin » Ymax  dmin , ANd dnax With the correspond- 
ing values of N, O in Figure 3, for the selection 
of the optimal operating conditions NV, O for the 
