Modern manufacturers demand high-performance, low costs and sustainability from design engineers. The challenge is to establish confidence in computational models using the minimum number of physical tests, precisely targeted to yield accurate, comprehensive and relevant data at minimum cost.
Model validation is a crucial step in establishing confidence in the computer simulations of how a device will behave. Research led by Professor Eann Patterson at the University of Liverpool is at the cutting-edge of research contributing to internationally-acceptable model validation methodologies.
Eann’s research group at Liverpool are contributing to high quality, industrially applicable digital models to meet design and safety demands. The work quantifies and reduces the uncertainty of strain measurement systems, which are used to test load-bearing structures of aircrafts or components for fission reactors, for example.
The group has developed efficient methods for quantitatively comparing computationally predicted strain fields of engineered components, with those measured in real physical tests. This was achieved by innovatively repurposing image decomposition techniques, originating from fingerprint and iris recognition. The developed model validation and calibration techniques can benefit a range of industries, including aerospace, automotive and nuclear.
Working in partnership
The team in Liverpool has worked with national labs including EMPA in Switzerland, and partners from the aerospace, automotive and energy industries through a series of EU-funded projects led by Patterson such as ADVISE (2008-2011), VANESSA (2012-2014) and MOTIVATE.
The efficient approach to compare information-rich data fields has spun-out new research partnerships. For example, collaboration with major aerospace manufacturer Airbus and the United States Air Force (USAF) is applying the research to quality assurance and in-service inspection of safety-critical components.
Outputs and outcomes
The new techniques developed by the Liverpool team have been incorporated to a pre-standard: Validation of computational solid mechanics models supported by the European Committee for Standardisation (CEN). CEN recommendations are being embedded into the internal procedures for validation of computational mechanics models by the collaborators.
The research has contributed to the development of ‘digital twin’ technology worldwide, where an electronic replica of an entire project is created to compare to real-world developments. This can lead to shortened development times, increased operability, reliability and safety through reducing the risk of unexpected failures, as well as reduced costs by minimising the energy embedded in the manufactured design.
The research is also influencing other fields, such as the integrated use of tests and simulations in fields as diverse as predictive toxicology to fusion energy.
Computer model validation and calibration techniques can reduce costs and improve reliability, safety and sustainability, benefitting industries from aerospace and automotive to nuclear.Professor Eann Patterson
Back to: School of Engineering