Practical results from a stochastic 3-dimensional multiphase flow solver

Uncertainty quantification (UQ) of multiphase flow systems is a developing field. While some work has been done to generate useful data regarding these flows, finding the best methodologies to efficiently do so is required. Due to the complicated nature of multiphase systems it can be difficult to collect experimental results and expensive to run simulations. To more fully understand the impact of input uncertainty on the outputs and physics in general, it is helpful to build up useful statistical data. The multiUQ project is a novel, intrusive stochastic multiphase flow program utilizing Navier-Stokes and a conservative level set approach with polynomial chaos for random variables. This code is now being developed in 3-D, providing stochastic results of real-world situations. Implementing a density-decoupled pressure solution method, the solver can now utilize more advanced approaches to solving the pressure Poisson equation, such as the multigrid method found in the HYPRE package. We present results of a stochastic falling droplet and atomizing jet, utilizing Sobol indices to determine the interactions which most impact system variability.