My current research is primarily concerned with the development of a computationally inexpensive turbulent multiphase simulation tool with application to turbulent liquid spray systems with complex geometries.

Current attempts in the literature, using either DNS (Direct Numerical Simulation) or LES (Large Eddy Simulation) are mainly limited to relatively simple geometries, and low Reynolds number flows.

The current work here is to simulate the motions of a turbulent spray by coupling the Lagrangian PDF (probability density function) formulation by Pope (1994) with a modified BBO (Basset Boussinesq Oseen) equation for each constituent droplet of the spray. Lagrangian PDF techniques are well known to provide accurate, and computationally inexpensive approximations to complex turbulent flow fields (Pope (1994).

The current technique also includes simulation of droplet deformation, breakup (primary and secondary), and appropriate evaporation models.

Although turbulent liquid sprays occur in a wide variety of environmental and industrial settings, the current work is concerned with providing a fundamental understanding of liquid fuel spray formation in rocket engines. The other objective is to develop the computational and analytical models for accurately predicting the fuel spray formation including deformation, breakup, and evaporation of droplets.

The Simplex droplet atomizer modeled in the present study (shown below) was also used by Kvasnak, Ahmadi, and Schmidt (in press), and McDonell & Samuelsen (1995).

The next two figures represent a two-dimensional sample of the instantaneous fluid velocity vectors and instantaneous streamwise velocity, respectively.

For the calculation of a continuous spray of nondeforming droplets, a qualitative comparison of the current simulation with results by McDonell & Samuelsen (1995) may be seen in the next figure. Good qualitative agreement in both spray angle and droplet distribution is shown for an ensemble of 10,000 spherical droplets.