iWeld Presented at ECNDT 2026
The iWeld consortium presented a range of research activities at the 14th European Conference on Non-Destructive Testing (ECNDT 2026), highlighting recent advances in intelligent weld inspection, ultrasonic modelling and data-driven characterisation of complex materials. The conference provided an opportunity to engage with the wider NDT community and showcase developments from several work packages of the project.
Presentations included:
- Ultrasonic characterisation of grain orientations, elastic tensor and geometry of thick-section welds using deep learning, demonstrating machine-learning approaches for material characterisation and weld inversion.
- Characterization of Microstructured Steels: from Bulk Waves to Guided Waves Radiation, presenting recent developments in understanding ultrasonic wave propagation in complex engineering materials.
- Addressing Uncertainties in Material Parameters for Nuclear Weld Imaging by the Development of an Adaptive Algorithm, focusing on adaptive imaging methods capable of improving defect detection in the presence of uncertain material information.
- Physics-Based Synthetic Data Generation for Non-Destructive Testing Applications, exploring the use of simulation and synthetic datasets to support the development and validation of advanced inspection approaches.
- Benefits of New Finite Elements Models Available for Ultrasonic NDE Simulations, highlighting new modelling capabilities for wave propagation and NDT simulation of complex structures.
