Prediction of single-point temperature statistics in a half-heated grid flow

A single-point probability density function is developed using a finite-difference numerical method in order to describe temperature statistics in the turbulence downstream from a half-heated biplane grid. The thermal field is decomposed into three distinct regions by analytical conditioning, and existing closure approximations are applied to each region. Results show an excellent agreement with the measured probability density functions, and with the four lowest order moments of temperature, over most of the flow cross section, except near the cold boundary. These results support the validity of the gradient transport model for turbulent transport of conditioned probability density functions and the molecular mixing closure approximation applied to the fluid in the thermal mixing layer.