next up previous contents
Next: The Enhanced Collision Rules Up: Development of the Lattice Previous: Lattice Boltzmann Model for

The Linear Collision Operator

The use of the above lattice Boltzmann equation removes the statistical noise from the lattice simulations. The collision operator tex2html_wrap_inline13945 still depends on the tex2html_wrap_inline13947 Boolean input and output states, where m=b for a model with b links and no `rest-particles' and m=b+1 for a model with rest-particles. For a Boltzmann simulation based on the two-dimensional FHP-III model tex2html_wrap_inline13955 is a tex2html_wrap_inline13957 matrix, for a face-centred-hypercubic (FCHC) model, used in three-dimensional simulations, tex2html_wrap_inline13955 is a tex2html_wrap_inline13911 matrix. Clearly the computational requirements for handling such a matrix can be limiting. The size of the matrix can be greatly reduced [40] by expanding the distribution function

equation1746

where tex2html_wrap_inline13963 is the equilibrium value of the distribution function and tex2html_wrap_inline13965 is the non-equilibrium part. Expanding the collision operator about the equilibrium distribution gives

equation1751

where

equation1755

Thus, since tex2html_wrap_inline13967 , the lattice Boltzmann equation can be written [40]

equation1763

Here tex2html_wrap_inline13975 is the linearised collision matrix and is an tex2html_wrap_inline13977 matrix, a considerable reduction from the tex2html_wrap_inline13979 matrix tex2html_wrap_inline13955 . The value of the individual elements of tex2html_wrap_inline13975 still depends on the form of the lattice gas collision rules.



James Buick
Tue Mar 17 17:29:36 GMT 1998