Science-based model for particle formation from novel fuels

Abstract

With the advent of petascale high-performance computing platforms, realistic multiscale modeling can be constructed to incorporate atomic-scale (molecular) information into macroscopic predictions of engineering systems. The overriding theme of the work presented in this paper is developing a multiscale modeling approach for soot formulation where atomistic data is integrated into macroscopic simulations. The prediction of soot formation remains arguably one of the most challenging subjects in combustion science, having an influence over a wide range of applications ranging from combustion efficiency to reducing emissions to slow global warming, to improved heat transfer designs in industrial settings, to predicting the radiation heat transfer from large scale fires. Starting from the fuel structures the new multiscale simulations reveals how chemical changes and transformation can propagate upward in scale to help define the function of the particle structures. In particular, the fuel structure influences the morphology of the nanoparticles, which in turn is critical in determining the overall growth and agglomeration behavior. These simulations make use of a newly proposed combination of molecular dynamics and kinetic Monte Carlo methodologies that will include both chemical reactions and agglomeration processes. The main strength of this approach is the ability to use important atomic-scale information directly into large scale description of the macroscopic phenomena.