Component shape optimisation for enhanced non-destructive testing

• Published in: Materials & design 2020-10, Vol.195, p.109041, Article 109041
• Main Authors: Singh, J., Mulholland, A.J., Tant, K.M.M., Stratoudaki, T., Curtis, A., Ijomah, W., Windmill, J.F.C.
• Format: Article
• Language: English
• Subjects: Design; Level set method; Non-destructive testing; Shape optimisation; Ultrasound
• ISSN: 0264-1275; 1873-4197
• DOI: 10.1016/j.matdes.2020.109041

Determining an optimal component design is fundamental to many applications including many of those in the aerospace, automotive and civil engineering industries. Previously, the compatibility of the component with future non-destructive testing (NDT) requirements has not been considered at the design stage. As a result, NDT operators are often challenged to reliably inspect components with complex geometries. In this paper, a framework is proposed for the optimisation of a component's shape to maximise the sensitivity and coverage of an interrogating ultrasonic wave, thus enhancing the ability to non-destructively image defects. The design of beam cross-sections are optimised using both a low-dimensional parameterisation with a genetic algorithm, and the level set method which enables more degrees of freedom in the component shape parameterisation. Images of flaws computed using the total focusing method show an 8 dB improvement in the signal to noise ratio for the optimised component as well as a five-fold improvement in the estimate of flaw size. These results show that the NDT-optimised design of components can provide significant improvements in flaw imaging. This in turn assists in extending the lifespan of in-service components and indeed their remanufacturability, which is both environmentally and economically advantageous. [Display omitted] •A design-for-testing framework for ultrasonic non-destructive testing suitability.•A novel implementation of the level set method based on ultrasonic coverage maps.•Optimised shapes lead to improved flaw detection.•Optimised shapes lead to more accurate estimates of defect size.