Document Type


Date of Degree

Spring 2013

Degree Name

MS (Master of Science)

Degree In

Occupational and Environmental Health

First Advisor

O'Shaughnessy, Patrick

First Committee Member

Peters, Thomas

Second Committee Member

Anthony, T R


Workplace aerosol sampling has been used to assess exposure to airborne materials that are known to cause adverse health effects in the respiratory system. Respirable cyclones are a common instrument used to monitor occupational exposures to respirable particles and are designed to have a penetration similar to the definition for the respirable fraction. However, deposited particles inside the walls of the cyclone may influence the penetration of cyclones. The aim of this study was to determine if there is a difference in collection efficiencies of a clean SKC 37-mm aluminum cyclone compared to a SKC 37-mm aluminum cyclone deposited with polydispersed dust.

Glass beads (Count Median Diameter CMD 3.3 µm, Geometric Standard Deviation GSD 1.7) were used to test a clean cyclone. The cyclone was then loaded by sampling with one of three dust types individually for three hours at concentrations of at least 3 mg/m3: Arizona Road Dust (CMD 1.04 µm, GSD 1.57), organic dust (CMD 2.90 µm, GSD 1.77), and titanium oxide (CMD 0.85 µm, GSD 1.28). After the cyclone was deposited with dust without cleaning, glass beads were used to retest the penetration. Particle penetration was measured using the Aerosol Particle Sizer (APS, TSI 3321). Particles depositing on the walls of the cyclone caused a shift in the penetration compared to clean samplers. When the cyclone was loaded with Arizona Road Dust, the penetration of particles increased as much as 5% at 3.5 µm. Depositing with Organic Dust increased particle penetration as much as 4% at 3.5 µm. Depositing did not occur with Titanium Oxide and did not significantly particle penetration.

Sampling with cyclones deposited with polydispersed particles can cause sampling errors by oversampling, and therefore overestimate the respirable concentration relative to a clean sampler. To counteract sampling errors from deposited particles would require the cyclone to be thoroughly dried and clean before sampling.


vii, 62 pages


Includes bibliographical references (pages 59-62).


Copyright 2013 William Leach