PhD Project opportunity with Liverpool University
Towards Data Driven Aerodynamic Models: Data Fusion of Experiment and Simulation (EPSRC CDT in Distributed Algorithms)
This studentship has been developed by the University of Liverpool in partnership with Aircraft Research Association Ltd.
The digital age with ubiquitous physics-based computational engineering tools, such as computational fluid dynamics (CFD), machine learning algorithms and ever-increasing computing power, helped accelerate the development of novel technologies deployed in the civil transport sector, as well as in defence and security, to meet the most demanding economic, environmental and societal challenges. One such example is when multidisciplinary CFD analysis is not only used routinely in the design of next-generation aircraft but also in the preparation of an experimental wind-tunnel test campaign to explore the parameter design space in a comprehensive and cost-effective manner, while ensuring the safe operation of the test. Notwithstanding, experimental studies, having greatly benefited from digitally enhanced data acquisition themselves, are essential in validating and fine-tuning numerical models, and even wholly predictive model building through machine learning, particularly when edge-of-the-envelope flow physics are involved.
Consequently, for the foreseeable future it is inconceivable that high-performance aircraft (and many other game-changing technologies) can be designed without physical wind-tunnel testing. Indeed, practical numerical methods often lack either prediction accuracy or the capability to model some physical phenomena altogether, or both. On the other hand, physical wind-tunnel testing not only becomes expensive when rapid design changes are sought (for which numerical tools are better suited), but, just as numerical tools, are subject to various uncertainties stemming e.g. from wind-tunnel corrections to account for the effects of wall constraint and flow field modification due to the chosen measuring technique or the fundamental flow characteristics of the wind tunnel itself. Experiment and simulation will remain in a symbiotic relation to produce highest-quality data and also to optimise experimentation, even to the extent of potentially increasing the efficiency (i.e. optimised power consumption) of industrial wind-tunnel testing.
Hence, fusion of disparate data from disparate sources (including both experiment and simulation) is paramount, promising step-changes in prediction capability overall to improve high-value design. Future design paradigms will use trusted discipline-based data models with quantified confidence levels. This is the overarching aim of the project. Specifically, it is envisaged to first explore future algorithms, including AI surrogate models, for near real-time joint experimental/numerical data analysis, that is uncertainty-aware, robust and quantifiable, to inform and optimise a wind-tunnel campaign, including on-the-fly. Second, considering the vast amount of data that a high-fidelity CFD run and a fully instrumented wind-tunnel test can produce, particularly for unsteady flow simulations, the first objective calls for high parallelisation utilising future computing systems, such as those explored within this CDT.
Further information can be found here (link to external site).