This paper describes the design optimisation study used to aerodynamically optimise the fairings that cover the rear wheels of the Land Speed Record vehicle, BLOODHOUND SuperSonic Car (SSC). Initially, using a Design of Experiments approach, a series of Computational Fluid Dynamics simulations were performed on a set of parametric geometries, with the goal of identifying a fairing geometry that was aerodynamically optimised for the target speed of 1,000 mph. Several aerodynamic properties were considered when deciding what design objectives the fairings would be optimised to achieve; chief amongst these was the minimisation of aerodynamic drag. A parallel, finite-volume Navier–Stokes solver was used on unstructured meshes in order to simulate the complex aerodynamic behaviour of the flow around the vehicle’s rear wheel structure, which involved a rotating wheel, and shockwaves generated close to a supersonic rolling ground plane. It was found that the simple response surface fitting approach did not sufficiently capture the complexities of the optimisation objective function across the high-dimensional design space. As a result, a Nelder–Mead optimisation approach was implemented, coupled with Radial Basis Function design space interpolation to find the final optimised fairing design. This paper presents the results of the optimisation study as well as indicating the likely impact this optimisation will have on the ultimate top speed of this unique vehicle.

No abstract

The land speed record vehicle, Bloodhound, undertook testing at subsonic and low transonic speeds (up to Mach 0.8) at Hakskeen Pan, South Africa, during October and November of 2019. A decade of CFD-led aerodynamic design had been undertaken to produce a vehicle with the aim of minimised Mach number aerodynamic dependencies and minimised overall drag. This paper sets out and explains the measured pressure distributions with a focus on the highest speed run of Bloodhound up to a peak speed of 628 mile/h. It compares the measured aerodynamic performance with the various CFD model predictions used throughout the design process showing that, whilst localised discrepancies between CFD model and real behaviour exist, overall the Reynolds-averaged Navier–Stokes (RANS)-based CFD tools used to design the car did result in sufficiently accurate aerodynamic data to predict the overall vehicle performance to a high degree of accuracy. The work outlined in this paper, and the conclusions and recommendations drawn, form the basis for a future record attempt and the understanding of what will be required in principle to extend the World Land Speed Record to 1000 mile/h. It also provides guidance on how to effectively make use of RANS-based CFD modelling predictions for other complex, ground-interacting high-speed applications.

Computational fluid dynamics solvers were applied to the field of high-speed boat design. The lattice Boltzmann method was used to assess the water-phase of the flow around a number of high-speed hullform geometries, and was validated against empirical industry and literature data. A heave dynamics capability was developed to assess the heave equilibrium position of a high speed boat, showing close agreement with industry data. A mesh movement and evolutionary optimisation software was applied to the aero-dynamic optimisation of a high-speed catamaran using a Reynolds-averaged Navier-Stokes solver for modelling of the air phase of the flow.