Issue 78

C. Vendittozzi et alii, Fracture and Structural Integrity, 78 (2026) 519-540; DOI: 10.3221/IGF-ESIS.78.30

screened for dropouts, saturation artifacts, and spikes, then converted to normalized strain and high-pass filtered. Event-centered analyses based on Hann-window short-time Fourier transform (STFT), Thomson multitaper estimates, and ridge tracking revealed recurrent transient signatures and sensor-dependent frequency families in two representative datasets. The results support FBG-derived landing-event characterization under field conditions, while showing that synchronized avionics references, temperature compensation, and larger repeated datasets are required before validated Health and Usage Monitoring System (HUMS) or structural-integrity decisions can be made. K EYWORDS . Fiber Bragg grating (FBG), Landing gear, Helicopter, Data integrity screening, Health and Usage Monitoring System (HUMS), Time– frequency Analysis.

I NTRODUCTION

anding gear assemblies are routinely exposed to high transient loads and broadband vibration during touchdown, roll-out, and ground maneuvers. These mechanical interactions can accelerate wear, promote localized damage, and trigger costly inspections, especially when hard-landing conditions are suspected. Within a Health and Usage Monitoring System (HUMS) framework, the availability of direct, quantitative indicators of landing gear response is therefore essential for enabling condition-based maintenance decisions and reducing reliance on indirect parameters or pilot reports [1]. In rotorcraft and broader aeronautical contexts, load monitoring at the landing gear has been discussed as a means to automatically trigger post-event inspections following hard-landings and to guide the inspection toward the most stressed components [2]. To address these requirements, fiber Bragg grating (FBG) networks have emerged as a prominent candidate technology for onboard deployment, offering well-known advantages such as electromagnetic immunity, low mass, multiplexing capability, and compatibility with distributed measurement layouts [3, 4]. The feasibility of integrating FBG strain sensors onto operational landing assemblies for remote, real-time load monitoring has been successfully demonstrated under representative loading conditions within laboratory environments [5]. Recent investigations have shown that FBG-based systems implemented on aircraft main landing gears yield load-regression accuracies with an error margin below 4%, providing strong evidence of their suitability for operational environments [6]. Further developments combine FBG measurements with advanced modelling strategies—including Gaussian Process Regression—to estimate loads with high fidelity even during controlled drop-test scenarios [7], demonstrating performance that is comparable, or even superior, to conventional strain-gauge instrumentation. In parallel, applications of FBG sensors in reusable launch-vehicle landing systems have confirmed their reliability under high-impact conditions, highlighting the critical role of optimized sensor placement for maximizing monitoring effectiveness [8]. Comprehensive reviews of FBG technologies for aerospace structural health monitoring further consolidate their role as a mature solution due to their sensitivity, long-term stability, and integration potential within advanced composite structures [9, 10, 11]. Literature also documents the application of FBG sensors to specific subcomponents, such as composite drag struts, supporting both stress monitoring and structural assessment tasks [12]. More recent work continues to report load monitoring solutions targeted at operational scenarios, including main landing gear applications during take-off and landing [13]. Hard-landing detection and severity assessment remain important long-term objectives for landing-gear instrumentation. Experimental studies have shown that FBG networks mounted on demonstrators can capture impact-related structural responses suitable for hard-landing assessment methodologies [14]. Recent developments have further proposed integrated monitoring architectures using FBG sensor networks to record operational loads, aiming to discriminate landing conditions and identify hard-landing events [15, 16]. These studies collectively motivate the use of optical strain measurements as direct observers of the landing event, complementing indirect approaches based on avionics parameters. However, moving from raw high-rate wavelength measurements to reliable operational descriptors remains challenging. Touchdown and roll-out responses are transient and nonstationary, and field recordings may include dropouts, saturation artifacts, short discontinuities, drift, and sensor position-dependent effects. Moreover, when synchronized avionics, weight-on-wheels signals, accelerometers, video records, or flight-log markers are not available, event identification must be treated as response-based segmentation rather L

520