Instructor: Malgorzata Peszynska, mpesz@math.oregonstate.edu.
Course announcement F25 MTH 654/9 Analysis and Approximation of Flow PDEs. MWF 9:00-9:50.
https://sites.science.oregonstate.edu/~mpesz/teaching/654_F25/

This course will introduce the stationary Stokes PDE system for viscous incompressible flow, its analysis and its finite element approximation. The model has the saddle point structure -\mu \Delta u + \nabla p=f, \nabla \cdot u = g. After variational formulation and numerical discretization, the model reads AU + B^T P =F, B U=G which is indefinite. Unlike many other elliptic PDEs which feature coercivity in Hilbert spaces (and positive definite linear systems), the Stokes system exploits functional analysis setting with Banach-Necas-Babuska conditions which derive from the Open Mapping and Closed Range theorems. The main theoretical tool is the generalization of coercivity to the inf-sup condition (also known as the LBB or Ladyzhenskaya-Babuska-Brezzi); these are verified in the continuous and discrete settings independently. In particular, the well-known MAC scheme will be interpreted as a lowest order mixed finite element approximation; the stability and accuracy of other CFD (Computational Fluid Dynamics) algorithms will be also discussed. As time permits, we will also explore time dependent and nonlinear (Navier-Stokes) problems, other saddle point problems with their Lagrangians, as well as appropriate iterative linear solvers.

The class will start with a gentle introduction to the modeling, analysis and computational aspects of fluids/CFD. Class notes and links to monographs available online will be provided. There will be no exams; theoretical and practical aspects of the material will be explored in assignments. Students interested in bona fide realistic simulations will use public domain libraries in C++ and/or MATLAB. Those interested in this class for enrichment rather than for a letter grade are welcome to discuss their options with me.

Required: solid background in differential equations, real variables, linear algebra, and experience with numerical methods. This class can be taken independently of any other courses MTH 654-5-6/9.

Thanks for your interest. I will be glad to discuss the class with you!