next up previous contents
Next: The Single Relaxation Model Up: The Boltzmann Equation Previous: Boltzmann's H-Theorem

The Chapman-Enskog Method

The Navier-Stokes equation can be derived from the Boltzmann equation using a Chapman-Enskog expansion. Only a brief outline of the method will be given here since the same procedure can be applied, in a much more straightforward manner, to the simplified lattice Boltzmann equation in which we are mainly interested. A full derivation of the Navier-Stokes equation using the Chapman-Enskog expansion will be given in section 4.2.3 for the lattice Boltzmann equation. Details of the Chapman-Enskog method for the classical Boltzmann equation can be found in [11].

The Chapman Enskog expansion parameter is the Knudsen number, tex2html_wrap_inline13299 , defined as

equation657

where tex2html_wrap_inline12727 is the mean free path of the molecules and l is a typical macroscopic length. The derived equations will only be valid if the Knudsen number is small. By analysing the time and length scales involved in the Boltzmann equation [13], tex2html_wrap_inline13299 can be introduced into the Boltzmann equation:

  equation662

Setting

equation671

we look for solutions of equation (2.37) such that

  equation677

The zeroth order them tex2html_wrap_inline13313 is taken to be the local Maxwell-Boltzmann distribution tex2html_wrap_inline13269 and tex2html_wrap_inline13317 , for tex2html_wrap_inline13319 , are chosen so they have no contribution to the moments expressed in equation (2.39). The first-order solution can be found by considering tex2html_wrap_inline13321 which gives the Euler equation of section 2.1.1 [11, 13]. The second-order solution, found by considering tex2html_wrap_inline13323 , requires a knowledge of the collision operator and can be shown [11] to give the Navier-Stokes equation when the binary collision function is used.


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