PyFR: Towards industrial adoption of scale-resolving simulations

Abstract: High-order numerical methods for unstructured grids combine the superior accuracy of high-order spectral or finite difference methods with the geometrical flexibility of low-order finite volume or finite element schemes. The Flux Reconstruction (FR) approach unifies various high-order schemes for unstructured grids within a single framework. Additionally, the FR approach exhibits a significant degree of element locality and is thus able to run efficiently on modern many-core hardware platforms, such as graphics processing units. Consequently, FR offers a promising route to performing affordable, and hence industrially relevant, scale-resolving simulations of hitherto intractable unsteady flows within the vicinity of real-world engineering geometries. In this talk I will review the current capabilities of PyFR, a high-order accurate computational fluids dynamics solver based on the FR approach. I will also present scale-resolving simulation results from a range of test cases which showcase the accuracy, scalability, and performance of PyFR.

Bio: Freddie Witherden studied Physics with Theoretical Physics at Imperial College London between 2008–2012 earning an MSci degree with first class honours. In September of 2012 Freddie started a PhD in computational fluid dynamics in the department of Aeronautics at Imperial College London under the supervision of Dr Peter Vincent and graduated in December 2015. He is currently an Assistant Professor of Ocean Engineering at Texas A&M University. He is also a fellow of the Royal Aeronautical Society and was named in the 2020 Forbes "30 under 30" list. Freddie's main research interests are in the development of simple and efficient high-order methods for modern CPU/GPU platforms. The primary vehicle for this is the flux reconstruction (FR) approach proposed by Huynh at NASA Glenn in 2007. Using the FR approach, it is possible to recover both discontinuous Galerkin and spectral difference schemes along with various new and novel methods. Furthermore, the approach is also extremely well suited to the requirements of modern many-core architectures.