Mainstream cigarette smoke generated using a Walton smoking machine and Kentucky 2R1 research cigarettes was studied. Results ~howed that puff volume and total particulate matter were consistent after the first puff, with average values of 35.6 cm' and 3.37 mg, respectively. The particle size distribution, measured with a multijet cascade impactor, was not related to butt length or relative humidity ( 1 95%), but was strongly dependent on the aging time. Based on simple monodisperse coagulation, the mass median aerodynamic diameter was calculated to be 0.45 pm at a dilution ratio of 21.7. Using a technique based on the dimensional change of collected droplet particles at various viewing angles of a scanning electron microscope, the count median diameter was

A cylindrical, Pyrex glass chamber with stable temperature control was evaluated to determine the residence time of aerosol particles within an enclosure, and to compare our results to the theoretical expression derived by Crump and Seinfeld (1981). Monodisperse polystyrene latex particles with diameters between 0.04 and 3 pm were used to determine the particle deposition rates under different water temperature gradients (AT,) between the top and bottom water layers of the chamber. The water layers and an insulating fiberglass jacket around the side wall helped to maintain a steady temperature gradient in the chamber and this was found to he very important in controlling particle deposition. Results indicated that the minimal deposition losses occurred at AT, = 0°C for particles between 0.2 and 0.3 pm, with a residence time as long as 38 h. Results also showed that particle deposition rate can be described by the theory of Crump and Seinfeld with an exponent n = 2.02 and a turbulence intensity k , = 0.0039/s at AT, = O0C, and n = 2.04 and k , = 0.020/s at AT, = 10°C gradient. Preliminary results also indicated that a slightly higher temperature at the bottom of the chamber (e.g., AT, = 5°C for 0.62-pm particles) might induce an upward thermophoretic force which would reduce particle settling. INTRODUCTIONThe deposition rate of aerosol particles in an enclosure depends on gravitational sedimentation, Brownian motion of the particles, and turbulent motion of the carrier gas. Gravitational settling controls the deposition rate of particles > 1 pm, whereas Brownian diffusion becomes the dominant factor when the particle size is < 0.1 pm. The turbulence of the gas medium in which the particles are suspended influences both the Brownian difision and the gravitational settling. As a result, the deposition of the particles becomes a complicated phenomena. Historically, the deposition behavior of aerosol particles in enclosures was studied during removal of radioactive aerosols formed in a reactor accident (Nuclear Energy Agency, 1979), and in an attempt to increase the efficiency of fine-powder production in a chemical

The aerodynamic behavior of aggregates consisting of uniform polystyrene latex (PSL) spheres and unaggregated cuboidal Natrojarosite particles in a TSI aerodynamic particle sizer (Model APS33B) has been studied. In initial tests, monodisperse PSL microspheres ranging from 0.3 to 7 y m in geometric diameter were generated from aqueous suspensions using a Lovelace nebulizer. APS33B responses for these uniform-sized particles showed multiple peaks. The major (primary) peak, which resulted from the smallest particle, corresponded to the unaggregated single spheres (singlets); the second, third, and fourth peaks were identified as doublets, triangular triplets, and tetrahedral quadruplets, respectively. Both doublets and triplets moved with their long axes in perpendicular (maximum drag) orientation to the flow direction in the APS33B. In contrast, the tetrahedral particles were isometric and had the same dynamic shape factor (drag resistance) for all three primary orientations. The particle Reynolds numbers (Re,) for these particles were calculated and ranged from 0.2 to 30 in the sensing volume of the APS33B detector (i.e., ultraStokesian conditions). Ultra-Stokesian drag forces for all three types of aggregates were, therefore, estimated and expressed as a function of an empirical factor (1 + a~e , b ) to the Stokesian drag force. The ultraStokesian drag of a Natrojarosite particle was measured in the range 20 < Re, < 50 and could be described with a similar expression. This approach facilitates the study of the dynamic behavior of nonspherical particles and yields new information about the characteristics of drag forces in the ultra-Stokesian regime.

The performance of a modified virtual impactor with a efficiency cuwe of the conventional type virtual impactor clean air core in the center of an aerosol flow is deand the ratio of clean air flow to the total flow rate. With scribed. With this feature, the modified virtual impactor a higher clean air flow, the 50% cutoff Stokes number has overcome the inherent disadvantage of conventional slightly increases and the efficiency curve has better designs which have fine particle contamination in the separation characteristics. Internal wall losses in the imcoarse particle fraction. Furthermore, the separation pactor are a function of particle size and reach a local efficiency of this impactor can be predicted from the maximum near the 50% cutoff size.

No abstract