S. Anand, Y.S. Mayya, M. Yu, M. Seipenbusch, G. Kasper

Journal of Aerosol Science, 2012, Volume 52, October 2012, 18-32

A comprehensive numerical study of the aerosol coagulation process with a constant nanoparticle source term in a well-mixed chamber has been carried out to better understand aerosol evolution under the condition of continuous injection (Seipenbusch et al., 2008). The simulations examine the evolution of the number concentration, size spectrum and mean size of the particles using the Fuchs kernel for the coagulation coefficient between the particles. The paper specifically focuses on the influence of key parameters such as the initial particle size, fractal dimension of the coagulated particles, particle injection rate and ventilation removal rate on the above aerosol metrics. It is found that the total number concentration attains a peak soon after nanoparticle emission starts, and then decreases monotonically if the ventilation is zero, or attains a steady-state limit in the presence of non-zero ventilation. Also, the number size distribution of this system gradually assumes a bimodal shape, with the larger mode attaining prominence more rapidly for fractal particles and at higher injection rates. Other significant results of the study include (i) a scaling relationship for the number concentration vs. time, (ii) establishment of an asymptotic decay law ∼t^⁎−0.4 (where t^⁎ is the scaled time) for the scaled number concentration under conditions of zero ventilation, and (iii) the existence of a critical ventilation removal rate at which the steady-state number concentration attains maximum value. The study also presents a simplified two-group model to understand the asymptotic behaviour analytically. The implications of these findings to work place environment are discussed.
A comparison with the experimental observations of Seipenbusch et al. (2008) shows good agreement with respect to the evolution of total number concentration, but only qualitative agreement in respect of the detailed size distribution. Possible reasons are indicated.