Mathematical Modelling and Simulations
Mathematical Modelling and Simulation focuses on developing advanced computational techniques to understand the nature of complex engineering systems and provide digital solutions to accelerate the engineering process and enhance the performance of the end product. Find out more below.
This research area covers a wide range of topics and is underpinned by 5 cross-cutting themes, including:
1. Multi-physics, multi-scale and multi-fidelity modelling
Understanding and optimising modern engineering systems requires integration of different physical phenomena, for example the electromagnetic excitation of an electric motor’s casing and the sound radiated as a result of this. Such phenomena can take place at different spatial and temporal scales. Effective integration of the relevant models and their systematic reduction into computationally efficient simulations is a core activity that allows our experts to tackle previously “impossible” engineering problems.
2. Machine learning
Application of machine learning techniques in engineering allows our researchers to leverage the wealth of experimental data generated in cutting-edge facilities such as our wind tunnels, anechoic chamber, tyre testing laboratory etc. Such data is used to directly train machine learning models of the relevant systems, or to develop hybrid data-based/physical models. In addition, machine learning research feeds into online/real-time parameter estimation and model tuning for complex systems.
3. Hardware-in-the-loop
Integration of real-time hardware testing with simulation models offers unique opportunities to validate and optimise system performance under realistic conditions.
4. Digital twins
Benefiting from developments in both machine learning and multi-physics modelling, our digital twin research activity looks at the development of efficient digital twins of various systems that can be used for monitoring, analysis and optimisation.
5. Computational Fluid Dynamics (CFD)
Although part of our broader multi-physics and multi-fidelity modelling portfolio, our cutting-edge CFD work stands on its own with a long track record of providing solutions to complex industrial problems such as reactive flows, multi-phase flows and magnetohydrodynamics. Our CFD work is supported by Rolls-Royce UTC.
Our research activities
Our research includes - but is not limited to - the following areas:
- Developing accurate CFD simulation methodologies for complex flows including single phase, multiphase and reacting systems
- Vehicle aerodynamics, including coupled vehicle dynamics and driver modelling for cross-wind stability and research to reduce drag and improve handling for road vehicles and cyclists
- Multi-physics modelling (electrical/NVH) of electric motors
- CFD modelling and simulation of road spray and its effect on driver assistance systems and vehicle surface contamination
- Continuation methods to assess the stability of highly non-linear systems
- Machine learning for quality control and digital twins
- Multi-body dynamics and multi-physics simulation of the whole vehicle and its subsystems (e.g., structural, friction and hydrodynamic lubrication of windscreen wipers)
- On-line/off-line tyre and vehicle model parameter identification (including off-road conditions)
- Turbine blade vibration modelling for high cycle fatigue
- Simulation of advanced combustion systems to reduce emissions and enable alternative fuel use
- Simulation of turbomachinery
- Aeroacoustic simulation to understand and reduce noise from aircraft.
