This is the final report of a three-year, Laboratory-Directed Research and Development (LDRD) project at the Los Alamos National Laboratory (LANL). In this LDRD, we have developed a highly compact and descriptive formalism that allows us to broach the theoretically formidable morass of inhomogeneous turbulence. Our formalism has two novel aspects: (a) an adaptation of helicity basis functions to represent an arbitrary incompressible channel flow and (b) the invocation of a hypothesis of random phase. A result of this compact formalism is that the mathematical description of inhomogeneous turbulence looks much like that of homogeneous turbulence--at the moment, the most rigorously explored terrain in turbulence research. As a result, we can explore the effect of boundaries on such important quantities as the gradients of mean flow, mean pressure, triple-velocity correlations and pressure velocity correlations, all of which vanish under the conventional, but artificial, assumption that the turbulence is statistically spatially uniform. Under suitable conditions, we have predicted that a mean flow gradient can develop even when none is initially present.
PDF
download
878987797
028-85220240
