Application of parallel object-oriented environment and toolkit (POET) to combustion problems

As an initial demonstration of the parallel object-oriented environment and toolkit (POET) the authors have investigated a probability density function (PDF) code for a jet flame diffusion problem. The PDF algorithm involves mostly Monte Carlo calculations and is highly amenable to an efficient parallel implementation. POET is used to partition the algorithmic portions of the code from the application-specific code. POET was designed to provide a mechanism that permits both the computational and the physical scientist to contribute their expertise to an integrated scalable application through the use of a well-defined object-oriented interface. Through this interface, direct integration of existing application codes is implemented without affecting the parallel computation algorithms. Thus, the combustion scientist need only be concerned with application-specific code. The results of interfacing a PDF jet flame diffusion model with POET have generated the first investigation of a turbulent nonpremixed flame with a full chemical mechanism. >