Modeling Coagulation, Aggregation and Gelation in High Volume Fraction Aerosols using Langevin Dynamics Simulations

RANGANATHAN GOPALAKRISHNAN, Zhibo Liu, Vikram Suresh, Zachary Perry, The University of Memphis

     Abstract Number: 28
     Working Group: Combustion

Abstract
Effect of volume-fraction and particle-particle hydrodynamic interactions influence particle coagulation in aerosols. Especially, at particle volume fractions exceeding 0.001%, runaway aggregation leads to aerosol gel formation as discovered experimentally (Dhaubhadel, R., et al. (2009). "Light Scattering Study of Aggregation Kinetics in Dense, Gelling Aerosols." Aerosol Science and Technology 43(11): 1053-1063).

In the first part of the presentation, a modeling study that uses Langevin Dynamics (LD) trajectory simulations of N mono-sized spherical particles in a periodic domain is discussed. The extended Kirkwood-Risemann approach (J. Fluid Mechanics 855, 535 (2018)) is invoked to compute particle-particle hydrodynamic interactions whose effect is parameterized as a function of the momentum Knudsen number (Kn). The results are summarized as a model for coagulation rate constant (βij) that depends on the diffusive Knudsen number (KnD) used in prior work to parameterize coagulation in the dilute regime (Aerosol Sci. Tech. 45, 1499 (2011)), Kn and particle volume-fraction ηv. In the absence of hydrodynamic interactions, it is observed that the coagulation rate constant in the continuum limit for mass transfer (KnD→0) is significantly enhanced by a factor of ~80 at ηv~0.3 due to particle crowding. While considering hydrodynamic interactions for ηv≥0.05, we use a screening distance around each particle that scales inversely with ηv beyond which the contribution of farther neighbors is neglected owing to the rapid decay of hydrodynamic interactions with distance. We also present new LD calculations of βij and elucidate the dependence of the same on Kn1 and the particle radii ratio θr for the coagulation of two particles in the dilute limit ηv→0. It is observed that the reduction of βij becomes significant as Kn1→0: at the lowest momentum Knudsen number considered (Kn1=0.1): βij is reduced by a factor of ~10 for equally sized particles (θr=0.5). At high KnD,Kn1, the particle size disparity is not significant, and it is seen that βij matches hard sphere predictions, indicating the insignificant contributions by hydrodynamic interactions. This study was recently published in in the J. Aerosol Science (https://doi.org/10.1016/j.jaerosci.2022.106001) and includes a series of animations of 2-particle simulations to illustrate the role of hydrodynamic interactions in particle coagulation. Computational results are summarized as regressions for convenient incorporation into particle/droplet growth sectional models.

In the second part of the presentation, current work that is yet to be published on modeling aggregation and gelation is discussed. Starting with N monomers, a Langevin Dynamics simulation that tracks particle aggregation until all the monomers are part of a single super-aggregate. Predictions of aerosol gel structure and morphology, their adhesion energy/strength are developed for comparison with appropriate experimental data. Secondary effects such as agglomerate rotational motion as well as translation-rotation coupling are discussed. Also presented is a modeling approach in the works to capture particle sticking and subsequent agglomeration without ad hoc assumptions. Considerations of experiment design are presented to enable model refinement.