The science focus in this project is on the study of the turbulent mixing of scalars with very low diffusivity, or very high Schmidt number, extended into regimes that have been difficult to reach. One specific target is to reach a Schmidt number of 512, which is comparable with that for salinity transport in the ocean, while retaining the flow properties that characterize high Reynolds number turbulence. A high Schmidt number is challenging for both experiment and computation because fluctuations arise at scales much smaller than those of the velocity field.
An innovative dual-grid, dual-communicator approach where scalar fields are computed using eighth-order compact finite difference methods in a computation-intensive code that is highly optimized using OpenACC on Titan is the key to a likely breakthrough. A second focus is the occurrence and topology of extreme events in scalar dissipation rate at high Reynolds number and moderately high Schmidt number. This work will likely provide definitive answers on questions such as Batchelor scaling, local isotropy and saturation of intermittency that have long remained obscure because of limited resolution and limited Schmidt number in previous works.
|Source||Hours||Start Date||End Date|