par Farina, Dario 
Président du jury Haut, Benoît
Promoteur Hendrick, Patrick
Co-Promoteur Iorio, Carlo Saverio
Publication Non publié, 2025-10-15
Président du jury Haut, Benoît
Promoteur Hendrick, Patrick
Co-Promoteur Iorio, Carlo Saverio
Publication Non publié, 2025-10-15
Thèse de doctorat
| Résumé : | In-flight ice formation poses significant challenges to aircraft safety and performance, necessitating advanced detection and protection systems. Ice accretion on critical aerodynamic surfaces can lead to increased drag, reduced lift, and compromised control, highlighting the urgent need for reliable detection technologies. Current solutions face limitations in accuracy, response time, and adaptability to different flight conditions.This thesis focuses on the development of a graphene-based smart ice detection system designed to detect ice presence, predict its formation, and measure ice thickness in real time. The system leverages PEDOT:PSS polymer sensors, whose electrical resistance varies upon contact with ice, and integrates artificial intelligence (AI) algorithms to enhance real-time detection capabilities. A key aspect of the research includes optimizing sensor manufacturing, developing robust electrical circuits, and creating software for signal interpretation.As part of the Graphene Flagship’s Spearhead Project “GICE”, this work aims to elevate the Technology Readiness Level (TRL) of graphene-enabled ice protection systems. The experimental activities were conducted at the Université libre de Bruxelles and SurfLabX’s iCORE Wind Tunnel, using a NACA 0012 airfoil as a testbed.The research methodology involved a series of wind tunnel tests under varying airflow conditions, including changes in airspeed, temperature, and water injection rate. Preliminary tests were also carried out in a temperature-controlled chamber to refine sensor performance. A total of over 200 controlled tests were performed, combining chamber trials and wind tunnel experiments. The graphene–PEDOT:PSS detector consistently demonstrated high reliability across these campaigns. In particular, the system achieved a detection accuracy of 97\%, with a precision and recall of 1.0 in controlled scenarios where ice formation was present. The average detection delay was less than 2 seconds after ice onset, meeting aeronautical requirements for real-time safety systems. Ice thickness measurements, calibrated through image-processing techniques, reached an average resolution of $\pm$40 $\mu$m, which is well within the expected tolerance for aviation ice detection applications. These quantitative results confirm the robustness of the proposed solution and its potential for scaling to higher Technology Readiness Levels (TRL).In conclusion, this thesis provides a detailed analysis of graphene-based ice detection technology and presents a path toward enhancing flight safety and sustainability. By addressing the challenges of in-flight ice formation with a novel approach, this research contributes to the Graphene Flagship initiative to bring graphene technologies from the laboratory to real-world applications, aligning with Europe’s strategic goals for innovation and sustainability. |
