| Abstract: | The ability of the Model 3320 aerodynamic particle sizer (APS) to make accurate mass-weighted size distribution measurements was investigated. Significant errors were observed in APS size distribution measurements with measured mass median aerodynamic diameters (MMADs) as much as 17 times higher than from cascade impactor measurements. Analysis of APS correlated time-of-flight and light scattering data indicated that the MMAD distortions were due to a few anomalous large particle measurements (～0.1% of the total measurements) with surprisingly low scattered light. Computational fluid dynamics modeling indicated that these anomalous measurements were due to particles that deviated from the intended aerosol pathway and recirculated through the APS measurement volume at low velocities leading to erroneous large particle measurements. A technique for removing erroneous measurements based on correlated aerodynamic diameter and light scattering values is presented. When this technique was used, APS and cascade impactor size distribution measurements agreed well. |
