Checking the Calculations: How Do Engineers Trust Their Designs?
This lecture is face-to-face on the University of Liverpool's campus, exact venue TBC. It is suitable for anyone aged 16 or over with an interest in mathematics. Please book one ticket for each person attending. There will not be a live stream or a recording.
Engineers are constantly developing new designs and technologies for a safer and more sustainable future. With our growing understanding of the world and computational capabilities, computer simulations have transformed how we study engineering systems and environment around us. A simulation uses maths to show how an engineering system or a scientific phenomenon works in real life. It helps scientists and engineers understand and test what might happen under different circumstances. For example, cars driving over a bridge or how an airplane behaves during a storm. Computer simulations are enabling more prototypes to be explored and more tests to be performed virtually at a fraction of the time and cost.
It’s important to show how much we can trust a simulation and how good or bad its results are. This can be achieved through a validation process, which involves comparing simulation results with physical measurements. We analyse similarities and differences by using statistical methods and conclude how good or bad the simulation is based on the data. It can help decision-makers to have confidence when considering simulation results and help to create safer, stronger, and more environmentally friendly designs for the future.
About the speaker
Dr Ksenija Dvurecenska is a lecturer in the School of Engineering at the University of Liverpool. She won the Chief Scientist’s Award (National Nuclear Laboratory) in 2021.
Her research focuses on developing validation methodologies to build confidence in predictions of computational models. She studies image decomposition (or orthogonal decomposition) techniques and statistical methods to compare outcomes from simulations, for example displacement fields, with full-field measurements. These techniques are applicable to a variety of disciplines and sectors where data can be represented as a map; currently, she is concentrating on applications in aerospace and nuclear engineering.
The challenges she is addressing are the standardisation of the validation methods, effective use of the measured data in the validation process and interpretation of the validation outcomes to support decisions based on the predictions of the computational models.