This project is focused on developing a new generation of multiphase flow models for porous medium systems, which have a very wide range of application throughout the geosciences, for engineered systems, and biological systems. The research team has developed a theory known as the thermodynamically constrained averaging theory (TCAT), which is consistent across length scales, constrained by the second law of thermodynamics, and hysteretic free. To complete the theory and render the work suitable for application, closure relations must be developed in a specific form, evaluated, and validated. This work will
use leadership scale computing, along with experimental approaches, to complete and validate a two-fluid flow TCAT model. The specific issues to be resolved are the scale at which the continuum methods commonly applied in practice rest on a solid foundation and how this length scale varies with respect to the
media properties. The remaining quantities needed to complete the model can be deduced from the state space of the system simulated at the microscale. Simulations will be performed using a highly ecient and scalable lattice-Boltzmann simulator. The simulation design will yield eciencies generally around 90%. The simulator state will be advanced on the GPUs and processing of the state space will be averaged to the desired macroscale quantities concurrently by the CPU cores, which will not degrade the performance. This approach is standard for the code base relied upon and drastically reduces data movement, while minimizing o-line processing.
|Source||Hours||Start Date||End Date|