SHIFT OF AEROSOL PENETRATION IN RESPIRABLE CYCLONE SAMPLERS

To investigate the effect of aerosol loading on cyclone performance, an ultrasonic atomizing nozzle was used to generate micrometer-sized particles as challenge aerosols. A 25-mCi Po-210 radioactive source neutralized the particles to the Boltzmann charge equilibrium. An aerodynamic particle sizer measured the particle size distributions and number concentrations upstream and downstream of the cyclones. A humidifying system controlled the relative humidity in the test chamber to study the effect of hygroscopic properties on aerosol penetration as a function of particle size. The change in cyclone collection efficiency curve was significant. For example, the aerosol penetration of 4 mu m particles dropped from 50% to about 30% for the nylon cyclone when challenged with small potassium sodium tartrate particles (count median diameter [CMD] of 3.5 mu m, geometric standard deviation [GSD] of 1.3, mass concentration of 4.6 mg/m(3)) for 3 hours. The deviation was less serious (from 50% to about 40%) for the multi-inlet cyclone, indicating that the lower inlet velocity may cause less aerosol deposition and reentrainment. The decrease in collection efficiency was almost invisible when the challenge aerosol was large (CMD of 7.4 mu m, GSD of 1.5, mass concentration of 15.1 mg/m3). The change in separation efficiency was significantly reduced when the relative humidity in the test chamber was increased from 10 to 60%, implying that the aerosol deposition and reentrainment on the side wall is strongly influenced by humidity. The use of a virtual cyclone might be the ultimate solution to prevent the shift of separation efficiency curve due to aerosol loading on the side wall of the cyclone, a phenomenon that inhibits the accuracy and precision of size-selective aerosol measurement