The Boeing hump – turbulent separated flow with sweep

Research on turbulent boundary layers relies on experiments, and increasingly as computational/virtual experiments.

High-fidelity computational fluid dynamics (HF-CFD) at relevant Reynolds numbers has become feasible during the last years.

Guided by the NASA roadmap for CFD, several key research directions are identified, including algorithms and implementation, but also canonical flow cases of immediate aeronautical interest.

We thus consider the recently by Boeing proposed "speed bump": the design mimicks the flow around an aircraft wing, and allows for deep analysis of the physics relevant for wing design incorporating complex phenomena such as surface curvature, re-laminarisation and re-transition and separation.

The goal of the present project is to study this fully 3D flow at high Reynolds number, necessitating significant advances in turbulence analysis methods.

In particular, we will employ novel adaptive meshes with data-driven error indicators, modal decompositions and uncertainty quantification to perform and analyse simulations of high reference quality.Physically, we will study the effect of pressure gradients and sweep on turbulent boundary layers.

Using modern spectral and modal decompositions, we assess for the first time independently the influence of surface curvature and crossflow on the near-wall turbulence.

We will also work towards wall models to identifiy possible bottlenecks in the modelling approach used for engineering predictions.