Skip to main content

European training Network in intelligent prognostics and Health mAnagement in Composite structurEs

Periodic Reporting for period 1 - ENHAnCE (European training Network in intelligent prognostics and Health mAnagement in Composite structurEs)

Período documentado: 2020-01-01 hasta 2021-12-31

The ENHAnCE project aims to train the next generation of scientists to produce a paradigm shift on the health management of composite structures by fusing ad-hoc sensing technologies and prognostics engineering within the structural system leading to a new concept of intelligent structures understood as cyber-physical systems.

The ENHAnCE’s technological opportunity lies in the transformation of composite structures from being physical to cyber-physical based on the integration of health monitoring data as a part of ‘on-board’ expert systems, including diagnosis and prognostics capabilities. The network will develop this technology from the design stage and will make the journey to technology development, thus enabling an effective knowledge and training transfer to industry and practitioners. ENHAnCE’s outputs will lead to an extended lifetime and optimal serviceability of composite structures with marked reduction of maintenance costs, which will directly impact the economy and society. The latter will mainly impact to composite-based industries like the aeronautics and wind-energy, since a big number of new aircrafts are made of composites. Also, the big majority of turbine blades (and those in offshore locations) are made of composite materials. In both industries, the maintenance is a key aspect for competitiveness and for sustainability.

The overall objectives are:
• Deliver innovative embedded sensors onboard able to identify damage signature in real-time (minimally invasive SHM skin) and manufacturing method to integrate them within composite plates;
• Formulate novel mathematical and simulation tools for analyzing the interaction of sensor’s signals with damage in composites;
• Creation of real-time self-adaptive prognostics algorithms using integrated sensor’s data;
• Development of a Cyber-Physical structural information system using post-prognostics information based on the Plausible Petri net (PPN) paradigm.

The research and technology developed in this project will contribute on making these key materials more sustainable and competitive.
From a generic perspective, the project is progressing accordingly to what planned irreespective of the circumstances of the COVID pandemic, which precisely began in the project's M3, just at the beginning of the action. This situation has made us adapt a number of management and working methodologies (meetings, some training events, etc.), but with no loss of any objective, milestiones, training, nor research activities of the project.
As an overview, since the beginning of the project (M1) until M24, 22 out of 39* deliverables have been submitted, 5 out of 12* milestones achieved, and 4 out of 11* training weeks completed, where the symbol (*) is used to indicate that these values correspond to total amounts of the whole project (namely, from M1 to M42).

From a technical & research perspective, the work performed from M1 to M24 can be schematically described as follows:

-Method for evaluation and verification of embedded sensor’s health state under high tensile loads (WP2; Paper submitted for publication);
-Mechanical analysis of the influence of manufacturing processes on the capabilities of integrated SHM of smart composites during their lifespan (WP2; Paper submitted for publication);
-A methodology for determining an optimal experimental manufacturing process on survivability of embedded sensors in loaded thermoplastic composites and influence of the cabling system (WP2; Paper submitted for publication);
-Numerical computing tool to optimally investigate sensor's signal interaction with composite damage, including non-linear effects (WP3; Paper submitted for publication);
-Investigation on prognostics signatures for efficient damage prediction in composite materials under fatigue multi-damage modes (WP4; Paper submitted for publication);
-Development of Bayesian filtering algorithms for crack prediction in composites under on-line monitoring data, along with for delamination shape evolution prediction (WP4; paper published in "Structural Control and Health Monitoring");
-Novel Bayesian Neural Network training method to effectivelly account for the uncertainty, with special application to the uncertainty in the fatigue of composite structures (WP5; paper published in "Engineering Applications in Artificial Intelligence"); [Elsevier]);
-Novel Physics-enriched Machine learning method for hybrid damage diagnostics with quantified uncertainty (WP5; Paper submitted for publication);
-Novel complexity reduction technique for Petri net based decision making algorithms for maintenance of structures (WP5; paper published in "Reliability Engineering & Systems Safety" [Elsevier]);
-New self-adaptive decision-making method through Reinforcement Learning and Petri net modelling for structural maintenance (WP5; Paper under development);
-New management model for composite turbine blades with consideration of damage modes (WP5; Paper under development);
The outputs produced during the first reporting period, which have been synthetized in the last section, are keystones for scientific research beyond the state of the art. From a general perspective, the following items can be considered as the progress done beyond the state of the art:
-New manufacturing method for efficiently embed piezoelectric sensors within a thermoplastic composite laminate, along with its numerical and mechanical verification:
-Novel semi-analytical method for lamb-wave propagation and its interaction with non-linear damage, with the ability to be implemented within micro-processor for its high computational efficiency. The method has applications to composite materials and beyond;
-Novel intelligent prediction algorithms with effective quantification of the uncertainty and with a novel non-gradient-based training method;
-Novel intelligent self-adaptive management models to autonomously find the optimal maintenance policy;

The potential impacts expected are:
-Enhancement of the career perspectives and employability of researchers, and contribution to their skills development.
-Contribution to structuring doctoral/early-stage research training at the European level and to strengthening European innovation capacity.
-Increase the competitiveness of the EU composite industry and make it converge towards the Industry 4.0 paradigm.

So far, up to M24, 3 scientific papers have been published in top-ranked peer-reviewed indexed journals for effectively disseminating the research results and 4 more have been sent and are awaiting their review to be published.
2 press releases have been launched and an informative video is being edited to be shared on the ENHAnCE's website and other selected channels (universities and beneficiary/partner institutions).
Participation in conferences has not been possible yet (M24) since most of them were cancelled because the Pandemic circumstances, except for on-line conferences. However, in-person participation in conference is foreseen in the next future.
Enhance Team in front of a composite piece
Enhance Team
Workshop of the 3th TW
Session of the 3th TW