Evaluation of two inversion techniques for retrieving health-related aerosol fractions from personal cascade impactor measurements

Personal cascade impactors are widely used in occupational aerosol exposure assessment. Appropriate algorithms must be used to determine the total particle size distribution from masses collected on the stages of the cascade impactor. Such algorithms should be regarded as integral components of the measurement system. When evaluating algorithms for reconstruction of size distributions from cascade impactor data, the eventual use of the size distribution must be considered. So, from an industrial hygiene perspective, an appropriate basis for comparison of given measurement systems is the accurate retrieval of the inhalable, thoracic, and respirable aerosol fractions as described by the new, internationally accepted, particle size- selective sampling conventions (which are expected to form the basis of future aerosol standards).This article compares two inversion routines in terms of their abilities to retrieve these aerosol mass fractions relative to the masses that would have been obtained using an ideal sampler that perfectly followed the sampling convention. The routines were used to invert measurements made with the Institute of Occupational Medicine personal inhalable dust spectrometer, a miniature cascade impactor that aspirates the inhalable aerosol fraction, and the results are presented graphically as contours of equal mass bias. The simplest algorithm, based on the a prior assumption of lognormality, appears to provide the best results.