Making composite processes more reliable thanks to real-time monitoring

The industrial challenges of the MONOCLE project

Resin infusion processes are now widely used to manufacture large and thick composite parts. However, these processes present several major industrial challenges:

• difficulty in controlling the process in real time
• high risk of non-compliance on unique or near-unique parts
• significant production costs in the event of a defect
• lack of decision-making tools for operators
• need to secure quality from the first production

In this context, Naval Group wanted to have solutions that would help them better control the infusion processes in order to guarantee the quality of composite parts while limiting the associated industrial and economic risks.

Technologies developed for the monitoring of composite processes

The MONOCLE project has made it possible to develop a real-time monitoring methodology  based on several complementary technological bricks:

• implementation of multi-physics sensors: cameras, thermocouples, flow meters, vacuum sensors and machine data to feed process monitoring models
• artificial intelligence data processing to identify the resin front in real time and detect any anomalies
• development of a high-fidelity numerical simulation model to reproduce the behavior of the material during the infusion process
• design of a simplified metamodel to provide predictions compatible with the dynamics of the process
• implementation of a digital twin to improve manufacturing management and monitoring
• development of an ergonomic human-machine interface that renders key indicators in a clear and actionable way for operators
• use of the OPC UA standard for the management and operation of industrial data flows

The operator interface has been developed in close collaboration with the teams of Naval Group, iQanto and Bureau Veritas Marine & Offshore in order to meet the needs of field users.

The project results and industrial impacts

The MONOCLE project demonstrated the feasibility of intelligent monitoring of composite processes by infusion for large and thick parts.

The work carried out made it possible to:

• provide operators with a decision support tool based on real-time data
• improve control of resin infusion processes
• limit production costs related to non-conformities
• anticipate potential anomalies during manufacturing
• to structure methodologies that can contribute to the future certification of composite processes
• pave the way for the use of the digital twin and artificial intelligence in the manufacture of complex composite parts

Thanks to the results obtained, the MONOCLE project contributes to the development of more reliable, safer and better controlled composite processes for industries using large parts.