Two Robust, Efficient, and optimally Accurate Algorithms for parameterized stochastic Navier-Stokes Flow Problems

When

April 4th, 2025 | 2:30 p.m. – 3:00 p.m.
Refreshments provided

Where

University Hall 4002

Speaker

Neethu Ravindran, UAB graduate student

Abstract

We present two accurate and efficient algorithms for solving a group of fluid flow problems based on the Navier-Stokes Equations (NSEs). Both methods use a time-stepping approach with the backward-Euler method and include a special technique called Ensemble Eddy Viscosity (EEV) for stability. The first method solves the equations together as a system (coupled approach), while the second method separates them using a projection-splitting technique with additional stabilization. We prove that both methods are stable and accurate, and we show that, with the right parameters, the projection method gives the same result as that from coupled method. We also combine these methods with Stochastic Collocation to study uncertainty in the solutions. Our numerical experiments confirm the accuracy of the methods and demonstrate that the projection-based approach is more efficient.

*All UAB students, faculty and guests are invited.