Comparison of the Coulter Multisizer and Aerodynamic Particle Sizer for obtaining the aerodynamic particle size of irregularly shaped dust

Abstract

When studying air quality it is often necessary to measure the aerodynamic size distribution of particles. True aerodynamic diameter must be measured using a gravitational settling method, which is impractical. Other methods exist that use other properties of the particles to estimate the aerodynamic diameter. One of the most common methods is the TSI Aerodynamic Particle Sizer, which measures the time of flight of the particles in an accelerated flow field. This time is related to the aerodynamic properties of the particles. Another method uses the Coulter Multisizer, which measures the volume of each particle that has been collected on a filter. An equivalent diameter for each particle can be found by assuming the particle is spherical, which can then be converted to an aerodynamic diameter using the density of the particles. Both methods have been shown to work well for spherical particles, but cannot account for the shape of the particles being tested, which can significantly increase the drag force on the particle. As a result the APS tends to undersize these particles, while the Coulter Method overestimates the aerodynamic size. This study compared the two methods of estimating particle size using a dust chamber where two different irregularly shaped dusts were supplied to three Coulter filters and the APS inlet. Results showed that the shape of the distributions from both the Coulter and APS methods were comparable, but the Coulter diameters were generally higher than those of the APS. This was consistent with what was expected, although there was significant variability in the Coulter results for one of the dusts that could not be explained. The APS also demonstrated significant problems with counting large "phantom" particles that did not actually exist. This problem was improved upon using a data "mask" provided by the manufacturer, but was not entirely corrected. Overall the two methods gave similar results indicating that a possible conversion factor could be developed to compare the two, but no simple method could be found using this data set. More tests need to be conducted to develop this conversion factor as well as comparisons to gravitational methods.