Scalings in Diffusion-Driven Reaction A+B→C : Numerical Simulations by Lattice BGK Models

Y. H. Qian,S. A. Orszag
DOI: https://doi.org/10.1007/bf02179978
1995-01-01
Journal of Statistical Physics
Abstract:We are interested in applying lattice BGK models to the diffusion-driven reactive system A + B --> C, which was investigated by Galfi and Racz with an asymptotic analysis and by Chopard and Droz with a cellular automaton model. The lattice BGK model is free from noise and flexible for various applications. We derive the general reaction-diffusion equations for the lattice BGK models under the assumption of local diffusive equilibrium. Two fourth-order terms are derived and verified by numerical simulations. The motivation of this study is to compare the lattice BGK results with existing results before we apply the models to more complicated systems. The scalings concern two exponents alpha and beta appearing in the production rate of C component R(x, t) similar to t(-beta)G(xt(-alpha)). We find the same values for alpha = 1/6 and beta = 2/3 as Galfi and Racz found at the long time limit. A Gaussian-like function for G is numerically obtained, which confirms a similar result of Galfi and Racz. On the one hand, when compared with the asymptotic analysis, lattice BGK models are easy to apply to cases where no analytic or asymptotic results exist; on the other hand, when compared with cellular automaton models, lattice BGK models are faster, simpler, and more accurate. The discrepancy of the results between the cellular automaton model and the lattice BGK models for the exponents comes from the role of the intrinsic fluctuation. Once the time and space correlation of stochastic stirring is given, we can incorporate a random fluctuating term in lattice BGK models. The Schlogl model is also tested, showing the ability of lattice BGK models for generating Turing patterns, which may stimulate further interesting investigations.
What problem does this paper attempt to address?