Numerical simulation of parallel-plate particle separator for estimation of charge distribution of PM2.5

A numerical simulation of an instrument that is used to measure the charging state of PM2.5 is conducted in order to clarify its measurement uncertainty and to improve its performance. The instrument, a parallel-plate particle separator (PPPS), is designed to classify aerosol particles according to their charging states and measure their quantities. The trajectories of submicron particles in the PPPS are numerically analyzed using the Lagrangian particle tracking method, taking into account the Brownian force and the electrostatic force. First, it is confirmed that the deterioration in the classification accuracy observed in the experiment is due to Brownian diffusion. The optimal condition that improves the accuracy is investigated through a parametric study by varying the balance of flow rates at the inlets, the geometry of the inlet and exit sections, and the applied voltage. It is found that decreasing the flow rate of the central inlet for aerosol or narrowing the central inlet improves the accuracy. The dependence of the accuracy on the flow rate is found to be in accordance with the experimental results. For charged particles, an optimum voltage that maximizes the classification accuracy is found. On the basis of the simulation results, we propose a method to determine the charge distribution of aerosol from the number of particles counted at each exit of the PPPS. In the test assuming aerosol in the air, the charge distribution determined from the number count at the exits is found to perfectly agree with the charge distribution specified at the inlet.