Errors in sampling particulate flows with a 10-mm Dorr-Oliver cyclone separator.

摘要

In this work, the efficiency of an isolated and mannequin-mounted 10mm Dorr-Oliver nylon cyclone as a particle preclassifier for respirable dust sampling was experimentally determined relative to an isokinetic sampler (sampling efficiency) and an inlet simulator sampler (penetration). A TSI Aerodynamic Particle Sizer (APS 3310) and a Marple six-stage impactor at different stream velocities (1.52 m/s, 2.43 m/s, 4.15 m/s), sampling flow rates (2.0 lpm, 1.7 lpm, 1.2 lpm) and inlet orientations ;The Sigmoid curve has been found to represent the efficiency (penetration and sampling efficiency) of the cyclone fairly well. The penetration of the isolated cyclone at a flow rate of 2.0 lpm, has been found to be much sharper than the BMRC or ACGIH (1985) respirable convention, whereas at a flow rate of 1.7 lpm, the penetration was found to follow the ACGIH (1985) respirable convention. The sampling efficiency of the cyclone was found to be a function of several different parameters such as stream velocity, sampling flow rate, inlet orientation, and the presence of a mannequin inside the wind tunnel. The orientation and velocity averaged sampling bias of the nylon cyclone was found to be ;Computer software CyDAP (Cyclone Design and Analysis of Performance) was developed to analyze the swirling flow inside the cyclone from the fundamental equations of multiphase flow. The collection efficiency model accounted for swirling gas-particle flow, particle entrainment, turbulent mixing, particle-wall interactions, boundary layer formation, particle leakage and particle drift toward the wall amongst others. Rather than attempting to analyze the cyclone flow using computational fluid dynamics techniques, CyDAP describes the different fluid mechanical phenomena inside the cyclone (as described above) with mathematical formulae and links them all to arrive at a unity of concepts from a series of interlinked phenomena. The collection efficiency and pressure drop of the nylon cyclone predicted using CyDAP matches the general trend of the experimental data. (Abstract shortened by UMI.).