Articles
Computational simulation of fluid flow, often referred to as computational fluid dynamics (CFD), plays a critical role in the design and manufacture of numerous complex systems, including civil/military aircraft, Formula One racing cars, and wind turbines. CFD technology allows engineers to understand complex flow patterns, and thus perform aerodynamic design, without ever building a physical prototype or firing up a wind tunnel.
Unfortunately, however, the capabilities of current-generation CFD software remain limited, and do not meet industrial needs in terms of efficiency and accuracy, especially when it comes to modelling unsteady vortex-dominated flows. Vortices play an important role in the design of various vehicles. For example, vortices forming above the delta wing of the Concorde aircraft were used to generate lift as it landed, and vortices over F1 cars are used to generate and control downforce. Unsteady vortex-dominated flows can also cause problems. For example, they are associated with a phenomenon called ‘flutter’: uncontrolled, violent, and ultimately dangerous flexing of aircraft wings. Unsteady-flow phenomena are also involved in the generation of aircraft noise – which is a hot topic given proposals for Heathrow expansion. So the ability to accurately simulate such flow phenomena is critical.
The objective is for a software tool that will help industries to move away from expensive physical prototyping and into the realm of digital prototyping
My research team at Imperial College London is developing a new generation of simulation tools for the efficient and accurate simulation of unsteady vortex-dominated flows. Our open-source software, PyFR (www.pyfr.org), is based on so-called ‘high-order’ algorithms that offer increased levels of accuracy. High accuracy is important when resolving unsteady vortical structures and/or acoustic phenomena. Our software can also make use of new ‘many-core’ hardware platforms, including GPU accelerators from Nvidia and AMD. Preliminary studies suggest that we can achieve order-of-magnitude increases in accuracy for the same computational cost as current-generation CFD technology.
Our end-goal is to provide industry with efficient, accurate, and reliable software tools for simulating hitherto intractable unsteady-flow phenomena. We hope these will ultimately improve the design of numerous systems, including civil aircraft (making them quieter), F1 racing cars (making them faster) and wind turbines (making them more efficient), while also reducing the need for physical prototyping.
To this end, the technology in PyFR is being commercialised via the Hyperflux project, which is funded by Innovate UK and ESPRC, and led by Imperial College London, the Centre for Modelling and Simulation, and Zenotech. Several organisations and industry associations have agreed to contribute ‘real-world’ test cases to help evaluate Hyperflux as it evolves. The ultimate objective is for Hyperflux to become a world-leading UK-based software tool that will help high-tech industries to move away from expensive physical prototyping and into the realm of digital prototyping.