Proceedings of the AAAR 23rd Annual Conference, October 4-8, 2004, Atlanta, Georgia. Mount Laurel, NJ: American Association for Aerosol Research, 2004 Oct; :35
Observations of the size of liquid droplets using the Aerodynamic Particle Sizer (APS, TSI, Inc.) and the Aerosizer (API, Inc. and TSI, Inc.) indicated that the measured size was significantly different from the aerodynamic diameter as measured by observing droplet settling velocity. The size shifts (Delta) were related to droplet aerodynamic diameter, viscosity and surface tension by the following empirical equation: Delta = a x diameter^b / (surface_tension^c x viscosity^e). The value of b was set to two. The values for a, c, and e were determined by a regression analysis of all the available data collected over several years with several models of each instrument. For the APS (Models 3300, 3320, 3321), the constants were: a = 1.22 x 10^-4; c = 0.5956; and e = 0.6916. For the Aerosizer (Models LD and DSP) the constants were: a = 4.061 x 10^-4; c = 0.9583; and e = 0.2516. The size shifts were initially attributed to droplet distortion Bartley et al. . This appears to be correct for the Aerosizer. However, for the APS, the situation was complicated by droplet deposition in the upper aerosol focusing nozzle. The nozzle Stokes diameter indicated that particles larger than about 5 micrometers can impact on the upper surface of the nozzle. Liquid built up in the nozzle, changing its shape and opening size, resulting in a velocity increase for particles passing through. When measuring particles, e.g. other droplets or solid particles, after liquid has deposited in the nozzle, the measured size shifted as much as 5 - 10%. After a few minutes this apparent size shift decreased by about half, but generally did not go away without cleaning of the nozzle. By cleaning the nozzle and observing the size shift immediately, the size shift due to droplet distortion was observed. The shift caused by nozzle loading then occurred usually within a few minutes at moderate concentrations of droplets larger than 5 micrometers. Thus, most of the measured size shifts for the APS as indicated by the equation above were caused by droplet loading in APS nozzle. Griffiths et al.  also documented significant size shifts for droplets observed with the APS. Their shifts were generally larger than in the present study and could be fitted using the above equations, though with different constants. While the above equations can be used to estimate the APS size shifts, these shifts may change with time and with liquid and, perhaps, solid particle loading.
Proceedings of the AAAR 23rd Annual Conference, October 4-8, 2004, Atlanta, Georgia