intelligent Weld inspection

Informations projet

iWeld

N° de convention de subvention: 101061359

Date de signature de la CE 7 Juin 2022

Date de début 1 Octobre 2022

Date de fin 31 Octobre 2026

Financé au titre de

• Euratom Research and Training Programme (EURATOM)

Coût total

€ 1 386 058,75

Contribution de l’UE

€ 1 161 136,00

Coordonné par

ELECTRICITE DE FRANCE
France

CORDIS fournit des liens vers les livrables publics et les publications des projets HORIZON.

Les liens vers les livrables et les publications des projets du 7e PC, ainsi que les liens vers certains types de résultats spécifiques tels que les jeux de données et les logiciels, sont récupérés dynamiquement sur OpenAIRE .

Livrables

Documents, reports (7)

External project communication plan

The project communication plan also covers the project's public website.

Simplified Weld Model

D2.1 delivers a simplified weld model, in order to overcome the limitation that full microstructure simulations are too costly to generate the number of datasets necessary to train an AI model. The deliverable describes a simplified procedure mapping a calculation of the vicinity of the thermal source to larger areas, according to defined welding conditions.

Weld uncertainty study for a chosen use case

D3.1 summarizes the weld uncertainty study for a chosen use case, determining the most relevant welding parameters and the expected confidence for associated tolerances, to answer the question of how accurate the orientation map needs to be for accurate imaging of ultrasound acquisition data.

Weld database

D1.1 delivers a database of industrially relevant weld shapes (V, X, K, J, etc.), welding techniques (TIG, SMAW) and positions (0, 90, etc.), including tolerances and typical uncertainties.

AI-based Weld Model

D2.2 constitutes an AI based weld model, which is delivered in the form of a public report and a partially public simulated database.

Demonstration of adaptive imaging on chosen samples

D4.1 presents experimental investigations of techniques developed on a particular set of samples

TRL Workshop

D5.1 provides a detailed assessment of the TRL level of the object of each task at the start of the project, as well as the targeted TRL level at the end of the project. The report summarizes the outcome of a TRL workshop to be held during the kick-off meeting.

Demonstrators, pilots, prototypes (1)

Inversion Methodology and demonstration

D3.2 demonstrates the developed inversion methodology and demonstrates it numerically on grain scale models. This deliverable is related to D4.1, which provides the practical demonstration on manufactured welds.

Publications

Peer reviewed articles (5)

Simulating ultrasonic inspections of welds with complex microstructure: a contribution to the IWELD European project

Auteurs: null null, Christophe Reboud, null null, Roberto Miorelli, null null, Nicolas Leymarie, null null, Frédéric Costes, null null, Chengdan Xue, null null, Edourad Demaldent
Publié dans: e-Journal of Nondestructive Testing, Numéro 30, 2025, ISSN 1435-4934
Éditeur: NDT.net GmbH & Co. KG
DOI: 10.58286/32341

Model-based deep neural network inversion for ultrasonic reconstruction of thick-section welds

Auteurs: Lucas Q. Machado, Thomas Blumensath, Vykintas Samaitis, Michael J.S. Lowe, Michał K. Kalkowski
Publié dans: NDT & E International, Numéro 163, 2026, ISSN 0963-8695
Éditeur: Elsevier BV
DOI: 10.1016/J.NDTEINT.2026.103775

Reactor performance, system reliability: instrumentation and control

Auteurs: Andreas Schumm, Mohamed Ben Chouikha, Bastien Poubeau
Publié dans: EPJ Nuclear Sciences & Technologies, Numéro 11, 2025, ISSN 2491-9292
Éditeur: EDP Sciences
DOI: 10.1051/EPJN/2025039

Simulating ultrasonic inspections of welds with complex microstructure: a contribution to the IWELD European project

Auteurs: null null, Christophe Reboud, null null, Roberto Miorelli, null null, Nicolas Leymarie, null null, Frédéric Costes, null null, Chengdan Xue, null null, Edourad Demaldent
Publié dans: e-Journal of Nondestructive Testing, Numéro 30, 2025, ISSN 1435-4934
Éditeur: NDT.net GmbH & Co. KG
DOI: 10.58286/32341

Improving defect detection of nuclear welds using adaptive imaging based on uncertain material parameters optimization

Auteurs: Ali Boukham, Jordan Barras, Maxance Marmonier, Nicolas Leymarie
Publié dans: Acta Acustica, Numéro 10, 2026, ISSN 2681-4617
Éditeur: EDP Sciences
DOI: 10.1051/AACUS/2026069

Conference proceedings (2)

Ultrasound propagation in materials with elongated grains and preferential orientation: parameter analysis and complex tensor determination

Auteurs: Machado, L.Q, Michal Kalkowski, Michael Lowe
Publié dans: ASNT Research Symposium, Numéro 32, 2024
Éditeur: ASNT

Ultrasonic characterisation of thick section welds assisted by AI and ray-based methods

Auteurs: Machado, L.Q, Thomas Blumensath, Michael Lowe, Michal Kalkowski
Publié dans: Annual Review of Progress in QNDE, Numéro 52, 2025
Éditeur: ASME