Abstract
Values were determined for the flow rate Q and calibration factor k which optimize the performance of the 10 mm cyclone in estimating respirable mass as defined by the British Medical Research Council (BMRC). To this end, flow dependent monodisperse cyclone collection efficiency data were analyzed as to reliability in predicting the collection of distributed dusts and were modelled mathematically. Furthermore, a simple criterion for evaluating cyclone performance was adopted. Performance maximization resulted in suggested values of 1.2 L/min and k=0.91 for BMRC dust concentration estimation. In order to achieve independence of the mathematical model, penetration measurements of the MSA sampling unit operated at 1.2 L/min (as well as 2.0 L/min) were carried out. Using results of these experiments, the bias between respirable and sampled dust concentrations was calculated in terms of geometric standard deviation and mass median diameter for a range of log-normally distributed dusts and is presented in graphical form. Sampling errors induced by variations in the flow rate from the optimizing values were also analyzed for dusts distributed log-normally. Preliminary results are also given for best estimating respirable dust concentration as defined by American Conference of Governmental Industrial Hygienists (ACGIH). The present work is expected to sharpen cyclone estimates of respirable dust concentrations as well as provide a simple method for fine-tuning other sampling devices.