Mitigating the Impact of Temperature Variations on Ultrasonic Guided Wave-Based Structural Health Monitoring through Variational Autoencoders.

Structural health monitoring (SHM) has become paramount for developing cheaper and more reliable maintenance policies. The advantages coming from adopting such process have turned out to be particularly evident when dealing with plated structures. In this context, state-of-the-art methods are based on exciting and acquiring ultrasonic-guided waves through a permanently installed sensor network. A baseline is registered when the structure is healthy, and newly acquired signals are compared to it to detect, localize, and quantify damage. To this purpose, the performance of traditional methods has been overcome by data-driven approaches, which allow processing a larger amount of data without losing diagnostic information. However, to date, no diagnostic method can deal with varying environmental and operational conditions (EOCs). This work aims to present a proof-of-concept that state-of-the-art machine learning methods can be used for reducing the impact of EOCs on the performance of damage diagnosis methods. Generative artificial intelligence was leveraged to mitigate the impact of temperature variations on ultrasonic guided wave-based SHM. Specifically, variational autoencoders and singular value decomposition were combined to learn the influence of temperature on guided waves. After training, the generative part of the algorithm was used to reconstruct signals at new unseen temperatures. Moreover, a refined version of the algorithm called forced variational autoencoder was introduced to further improve the reconstruction capabilities. The accuracy of the proposed framework was demonstrated against real measurements on a composite plate.

Physics-Informed Neural Networks for the Condition Monitoring of Rotating Shafts.

Condition monitoring of rotating shafts is essential for ensuring the reliability and optimal performance of machinery in diverse industries. In this context, as industrial systems become increasingly complex, the need for efficient data processing techniques is paramount. Deep learning has emerged as a dominant approach due to its capacity to capture intricate data patterns and relationships. However, a prevalent challenge lies in the black-box nature of many deep learning algorithms, which often operate without adhering to the underlying physical characteristics intrinsic to the studied phenomena. To address this limitation and enhance the fusion of data-driven methodologies with the fundamental physics of the system under study, this paper leverages physics-informed neural networks (PINNs). Specifically, a simple but realistic numerical case study of an extended Jeffcott rotor model, encompassing damping effects and anisotropic supports for a more comprehensive modelling, is considered. PINNs are used for the estimation of five parameters that characterize the health state of the system. These parameters encompass the radial and angular position of the static unbalance due to the disk installed on the shaft, the stiffness along the principal axes of elasticity, and the non-rotating damping coefficient. The estimation is conducted solely by exploiting the displacement signals from the centre of the disk and, to showcase the efficacy and precision provided by this novel methodology, various scenarios involving different constant rotational speeds are examined. Additionally, the impact of noisy input data is also taken into account within the analysis and the performance is compared to that of traditional optimization algorithms used for parameters estimation.

On the Evaluation of a Coupled Sequential Approach for Rotorcraft Landing Simulation.

Maximum loads acting on aircraft structures generally arise when the aircraft is undergoing some form of acceleration, such as during landing. Landing, especially when considering rotorcrafts, is thus crucial in determining the operational load spectrum, and accurate predictions on the actual health/load level of the rotorcraft structure cannot be achieved unless a database comprising the structural response in various landing conditions is available. An effective means to create a structural response database relies on the modeling and simulation of the items and phenomena of concern. The structural response to rotorcraft landing is an underrated topic in the open scientific literature, and tools for the landing event simulation are lacking. In the present work, a coupled sequential simulation strategy is proposed and experimentally verified. This approach divides the complex landing problem into two separate domains, namely a dynamic domain, which is ruled by a multibody model, and a structural domain, which relies on a finite element model (FEM). The dynamic analysis is performed first, calculating a set of intermediate parameters that are provided as input to the subsequent structural analysis. Two approaches are compared, using displacements and forces at specific airframe locations, respectively, as the link between the dynamic and structural domains.

Robust Identification of Strain Waves due to Low-Velocity Impact with Different Impactor Stiffness.

Low-velocity impacts represent a major concern for aeronautical structures, sometimes producing barely detectable damage that could severely hamper the aircraft safety, even with regards to metallic structures. For this reason, the development of an automated impact monitoring system is desired. From a passive monitoring perspective, any impact generates a strain wave that can be acquired using sensor networks; signal processing techniques allow for extracting features useful for impact identification, possibly in an automatic way. However, impact wave characteristics are related to the impactor stiffness; this presents a problem for the evaluation of an impact-related feature and for the development of an automatic approach to impact identification. This work discusses the problem of reducing the influence of the impactor stiffness on one of the features typically characterizing the impact event, i.e., the time of arrival (TOA). Two passive sensor networks composed of accelerometers and piezoelectric sensors are installed on two metallic specimens, consisting of an aluminum skin and a sandwich panel, with aluminum skins and NOMEXTM honeycomb core. The effect of different impactor stiffnesses is investigated by resorting to an impact hammer, equipped with different tips. Subsequently, a method for data processing is defined to obtain a feature insensitive to the impactor stiffness, and this method is applied to multiple impact signals for feature uncertainty evaluation.

Neutron shielding for a new projected proton therapy facility: A Geant4 simulation study.

In this work, we used the Monte Carlo-based Geant4 simulation toolkit to calculate the ambient dose equivalents due to the secondary neutron field produced in a new projected proton therapy facility. In particular the facility geometry was modeled in Geant4 based on the CAD design. Proton beams were originated with an energy of 250MeV in the gantry rooms with different angles with respect to the patient; a fixed 250MeV proton beam was also modeled. The ambient dose equivalent was calculated in several locations of interest inside and outside the facility, for different scenarios. The simulation results were compared qualitatively to previous work on an existing facility bearing some similarities with the design under study, showing that the ambient dose equivalent ranges obtained are reasonable. The ambient dose equivalents, calculated by means of the Geant4 simulation, were compared to the Australian regulatory limits and showed that the new facility will not pose health risks for the public or staff, with a maximum equivalent dose rate equal to 7.9mSv/y in the control rooms and maze exit areas and 1.3·10-1mSv/y close to the walls, outside the facility, under very conservative assumptions. This work represents the first neutron shielding verification analysis of a new projected proton therapy facility and, as such, it may serve as a new source of comparison and validation for the international community, besides confirming the viability of the project from a radioprotection point of view.