This paper describes recent progress made in the development and imple
mentation of a two-dimensional Monte Carlo/PDF/SPH particle method. Th
e approach, which is applicable to compressible turbulent Bows, incorp
orates elements of smoothed particle hydrodynamics to extract mean qua
ntities from the particles, including the mean pressure gradient. A ne
w and efficient algorithm based on Fourier-series expansions of the SP
H kernel is used to compute these means; for a simulation with N parti
cles the computational work scales purely as O(N). 4 thorough study of
numerical errors introduced by the finite series expansion is also pe
rformed, and results are presented which show that these errors scale
as expected and can he made negligibly small using modest computing re
quirements. The particle method has been used to simulate a variety of
2D flows, including a stationary turbulent plane wake, flows under th
e influence of simple body forces. and an unsteady flow featuring comp
ression/expansion waves and a pair of decaying vortices. The plane wak
e calculation includes comparisons with self-similar experimental data
and good agreement is obtained for the mean velocity and Reynolds str
ess profiles. (C) 1998 Academic Press.