“…Suppose, further, that one was interested in the diffusion-controlled mass rate of partial condensation of these molecules on a “cold” surface exposed to, or containing, this mixture. In the absence of an appreciable Ludwig−Soret contribution, this rate will often 33 be proportional to the triple integral
where D eff [ m ,σ,ε, T ( x , t )] is the pseudo-binary Fick diffusion coefficient through the prevailing background gas, given, say, by Chapman−Enskog theory, and the integral is carried out over all positive values of m , σ, and ε. Then, because the Chapman−Enskog collision integrals can usually be approximated as a power law (say, −0.17 exponent) in the relevant dimensionless temperature T /(ε eff / k B ), where ε eff ≈ (εε N 2 ) 1/2 , one finds that total local mass deposition rate, ṁ ‘ ‘( x , t ), will be approximately proportional to a particular mixed moment: M i , j , k of n ( m ,σ,ε; x , t ) with respect to m , σ, and ε, where i = 1, j = − 4 / 3 , and k = −0.057.…”