The digitalization era has introduced an abundance of data that can be harnessed to monitor and predict the health of structures. This paper presents a comprehensive framework for post-prognosis decision-making that utilizes deep reinforcement learning (DRL) to manage maintenance decisions on multi-component systems subject to imperfect repairs. The proposed framework integrates raw sensory data acquisition, feature extraction, prognostics, imperfect repair modeling, and decision-making. This integration considers all these tasks independent, promoting flexibility and paving the way for more advanced and adaptable maintenance solutions in real-world applications. The framework's effectiveness is demonstrated through a case study involving tension-tension fatigue experiments on open-hole aluminum coupons representing multiple dependent components, where the ability to make stochastic RUL estimations and schedule maintenance actions is evaluated. The results demonstrate that the framework can effectively extend the lifecycle of the system while accommodating uncertainties in maintenance actions. This work utilizes the Value of Information to choose the optimal times to acquire new data, resulting in computational efficiency and significant resource savings. Finally, it emphasizes the importance of decomposing uncertainty into epistemic and aleatoric to convert the total uncertainty into decision probabilities over the chosen actions, ensuring reliability and enhancing the interpretability of the DRL model.

Maintenance decisions often involve choosing between replacement and repair. The shortage of essential replacement parts has led to increased exploration of repair methodologies. However, repairs are often imperfect, leading to additional uncertainties in predicting the component's future condition. Existing approaches in the literature for modeling imperfect repairs struggle when repair dynamics are unknown requiring a large amount of data to be reliable. Furthermore, current methods are very task-specific, which limits the optimization of maintenance planning of varying components. This research addresses these challenges by conceptualizing imperfect repair effects as a stochastic increase in Remaining Useful Life (RUL). An existing deep learning model extracts prognostic-related features that can be utilized by any prognostic model to estimate RUL based on sensor data. Then, the proposed imperfect repair model predicts the RUL increase post-repair. This method offers three key benefits: (i) proactive post-repair assessment for improved maintenance, (ii) a data-driven repair model compatible with existing prognostic models, and (iii) flexibility in adapting to different repair techniques. Evaluation of the proposed model is conducted through tension-tension fatigue experiments on aerospace-grade aluminium specimens subject to imperfect repair. Results demonstrate the model's ability to accurately estimate the post-repair stochastic RUL increase.

In aircraft composite structures, impact-induced delamination poses a significant threat to their integrity, necessitating meticulous inspections to ensure reliable operation. However, monitoring delamination growth with existing nondestructive methods remains challenging due to the intricate nature of the damage mechanisms involved. This study introduces a novel approach by integrating guided waves (GWs) and electromechanical impedance (EMI) to achieve the prediction of remaining useful life (RUL) in woven-type carbon fiber-reinforced polymer (CFRP) plate-like structures subjected to compression fatigue conditions following a low-velocity impact. The novelty of this work lies in the fusion of GW and EMI techniques for the prediction of RUL, which is integrated into a comprehensive prognostic framework. Damage indicators (DIs) derived from GW and EMI measurements were first analyzed for their correlation with measured delamination growth and then used as inputs for prognostic models developed using deep neural networks. This approach significantly enhances the accuracy and reliability of RUL predictions as the proposed GW–EMI fusion models aim to harness the most effective predictions from each DI. An evaluation of the DIs revealed that GW–DIs achieved better accuracy on average across all cycles compared to EMI–DIs. Both fusion models demonstrated strong accuracy for individual samples, with Fusion Model 1 (RUL-fus-1) showing a 12% improvement and Fusion Model 2 (RUL-fus-2) showing a 24% improvement across all cycles on average. Notably, Fusion Model 2 exhibited the lowest error in the final cycles, with a 48% improvement in accuracy compared to the least successful model, demonstrating its potential for more precise prognosis through the integration of GW-DIs and EMI-DIs.

Designing health indicators (HIs) for aerospace composite structures that demonstrate their health comprehensively, including all types of damage that can be adaptively updated, is challenging, especially under complex conditions like impact and compression-fatigue loadings. This paper introduces a new AI-based approach to designing reliable HIs (fulfilling requirements—monotonicity, prognosability, and trendability—referred to as ’Fitness’) for single-stiffener composite panels under fatigue loading using acoustic emission sensors. It incorporates complete ensemble empirical mode decomposition with adaptive noise for feature extraction, semi-supervised base deep learner models made of long short-term memory layers for information fusion, and a semi-supervised paradigm to simulate labels inspired by the physics of progressive damage. In this way, nondifferentiable prognostic criteria are implicitly implemented into the learning process. Ensemble learning, especially using a semi-supervised network built with bidirectional long short-term memory, improves HI quality while reducing deep learning randomness. The Fitness function equation has been modified to provide a more trustworthy foundation for comparison and enhance the practical reliability of the standard in prognostics and health management. Ablation experiments are conducted, including variations in dataset division and leave-one-out cross-validation, confirming the generalizability of the approach.

Abstract: Remaining useful life predictions depend on the quality of health indicators (HIs) generated from condition monitoring sensors, evaluated by predefined prognostic metrics such as monotonicity, prognosability, and trendability. Constructing these HIs requires effective models capable of automatically selecting and fusing features from pertinent measurements, given the inherent noise in sensory data. While deep learning approaches have the potential to automatically extract features without the need for significant specialist knowledge, these features lack a clear (physical) interpretation. Furthermore, the evaluation metrics for HIs are nondifferentiable, limiting the application of supervised networks. This research aims to develop an intrinsically interpretable ANN, targeting qualified HIs with significantly lower complexity. A semi-supervised paradigm is employed, simulating labels inspired by the physics of progressive damage. This approach implicitly incorporates nondifferentiable criteria into the learning process. The architecture comprises additive and newly modified multiplicative layers that combine features to better represent the system’s characteristics. The developed multiplicative neurons are not restricted to pairwise actions, and they can also handle both division and multiplication. To extract a compact HI equation, making the model mathematically interpretable, the number of parameters is further reduced by discretizing the weights via a ternary set. This weight discretization simplifies the extracted equation while gently controlling the number of weights that should be overlooked. The developed methodology is specifically tailored to construct interpretable HIs for commercial turbofan engines, showcasing that the generated HIs are of high quality and interpretable.

Damage imaging plays a crucial role in structural health monitoring (SHM) systems for fast and efficient damage assessment. Delay-and-sum (DAS) beamforming is a widely used algorithm in non-destructive testing for damage imaging, but its effectiveness is often compromised by the use of sparse ultrasonic transducer arrays and the difficulty in detecting progressive delamination larger than the wavelength using guided wave-based methods under fatigue loading. Although X-ray imaging offers detailed assessments of progressive delamination, its application is still limited due to the need to interrupt fatigue loading cycles and its high operational cost. To this end, we propose a novel Damage Imaging framework that uses the fine-tuned Conditional Diffusion Model for SHM systems (DI-CDM). Leveraging the powerful image generation capabilities of diffusion models, the framework was fine-tuned by combining DAS beamforming images derived from ultrasonic sparse array data with X-ray images captured during fatigue loading cycles of the composite structures. The proposed approach can generate damage images that reveal the progression of delamination size in the fatigue loading process. The framework was validated through numerical simulations and experimental data from NASA datasets for composite structures, demonstrating its potential and effectiveness by applying diffusion models in SHM applications to enable fast, high-resolution damage imaging.

In recent years, prognostics gained attention in various industries by optimizing maintenance, boosting operational efficiency, and preventing costly downtime. Central to prognostics is the Remaining Useful Life (RUL), representing the critical time before system failure. Deep learning advancements facilitate RUL forecasting by extracting features from diverse data formats such as time series, images, or sequences thereof, in one, two, or three dimensions, respectively. Yet, predicting RUL from image sequences often relies heavily on resource-intensive techniques like digital image correlation, complicating data acquisition. To address challenges with high-dimensional data and unreliable models, this study introduces ISTRUST, an innovative Transformer-based architecture. ISTRUST (Interpretable Spatiotemporal TRansformer for Understanding STructures) tackles the dual challenges posed by high-dimensional data and the black-box nature of existing models. Leveraging Transformers’ attention mechanism, ISTRUST breaks down the spatiotemporal domain, effectively realizing interpretable RUL predictions under uncertainty using only sparse raw image sequences as input. Evaluated on fatigue-loaded composite samples showcasing crack propagation, ISTRUST interprets the relation between cracks and RUL via the attention mechanism. The results substantiate its capacity to interpret and clarify instances in which predictions may exhibit variability in accuracy. Through the attention mechanism, a strong correlation between the model’s spatiotemporal focus and the RUL predictions is established, making it, to the best of our knowledge, the first model to provide interpretable stochastic RUL predictions directly from sequential images of this nature.

Fatigue damage prognosis always requires a degradation model describing the damage evolution with time; thus, the prognostic performance highly depends on the selection of such a model. The best model should probably be case specific, calling for the fusion of multiple degradation models for a robust prognosis. In this context, this paper proposes a scheme of online fusing multiple models in a particle filter (PF)-based damage prognosis framework. First, each prognostic model has its process equation built through a physics-based or data-driven degradation model and has its measurement equation linking the damage state and the measurement. Second, each model is independently processed through one PF to provide one group of particles. Then, the particles from all models are adopted for remaining useful life prediction. Finally, the particles from each PF are fused with those from all the other PFs to improve their particle diversity, and consequently, to provide better estimation and prognostic performance. The feasibility and robustness of the proposed method are validated by an experimental study, where an aluminum lug structure subject to fatigue crack growth is monitored by a guided wave measurement system.

Continuous monitoring of spray velocity during the cold spray process would be desirable to support quality control, as spray velocity is the key process parameter determining the deposit quality. This study explores the feasibility of utilising Airborne Acoustic Emission (AAE) for real-time monitoring of spray velocity. Six spray tests were conducted, varying pressure and temperature to achieve different velocities. Optical means were used to measure velocity; while, the signal from the AAE was captured during deposition via a microphone. Features demonstrating a strong correlation with velocity were extracted from the acoustic signals. Both rule-based and machine learning models were employed to identify the moments where the nozzle was engaged with the substrate and diagnose the velocity. The results indicate that monitoring the spray velocity of the cold spray process using AAE is feasible.

RFID (Radio Frequency Identification) is commonly used to monitor goods along the supply chain. The feasibility of the application of a RFID tag to monitor CFRP components was demonstrated in a previous work by some of the authors. This work is aimed at evaluating how the integration of a RFID tag inside a CFRP laminate affects the quasi-static strength and the failure mode. Finite element modelling is used to design a specimen with an embedded tag, that may be representative of the tensile state of stress in a larger laminate. Tensile testing is performed both up to failure and by interrupting tests at different load levels in order to inspect the specimen by C-scan. Acoustic Emission (AE) and Digital Image Correlation (DIC) are used as additional non-destructive monitoring techniques. Some interrupted tests were selected to extract micrographic sections that illustrate damage development in the laminate. The extensive destructive and non-destructive characterization allowed to quantify the effect of embedding a RFID tag in the laminate in terms of strength decrease and failure mode.

Wind turbine blades carry the risk of impact damage during transportation, installation, and operation. Such impacts can cause levels of damage that can propagate throughout the structure compromising performance and safety. In this study, the effect of impact damage on fatigue damage propagation in test specimens representative of a spar cap-shear web adhesively-bonded connection of a wind turbine blade was investigated. In addition, the effectiveness of using acoustic emissions to detect early impact-induced fatigue damage was studied. Three impact tests with increasing levels of energy were investigated. The results showed that for an impact test with an average energy of 16.32 J, the fatigue damage accumulation process was not influenced by the size and location of the impact damage. But for impact tests with an average energy of 23.68 J and 32.13 J, greater crack density and accelerated de-lamination and de-bonding of the adhesive from the laminate could be seen in the impact zone. Acoustic emission was shown to identify the position of the damage zone for the higher energy impact tests. It was also effective in showing the progressive accumulation of fatigue damage in this zone during the fatigue test.

Acousto-ultrasonic composite transducers (AUCTs), comprising piezoceramic materials in a reinforced polymeric matrix, show promise for structural health monitoring in composite structures. Challenges arise when integrating AUCTs onto highly loaded thermoplastic composites, especially low-surface-energy materials like polyaryletherketone composites. To address this, the study explores the viability of attaching AUCTs to low-melting polyaryletherketone carbon fiber-reinforced thermoplastic composite structures using ultrasonic welding. This welding technique forms a joint where the interface material fuses with the AUCT embedment and the structure matrix, providing a reliable and automatable process. The investigation includes a comparative analysis of an ultrasonic welded joint with an external energy director and a reference AUCT system integrated using a vacuum bagging oven procedure. Results highlight the potential of AUCT configurations integrated by ultrasonic welding as an alternative solution, acknowledging challenges that persist for further development and increased reliability in structural health monitoring applications.

Developing comprehensive health indicators (HIs) for composite structures encompassing various damage types is challenging due to the stochastic nature of damage accumulation and uncertain events (like impact) during operation. This complexity is amplified when striving for HIs independent of historical data. This paper introduces an AI-driven approach, the Hilbert transform-convolutional neural network under a semi-supervised learning paradigm, to designing reliable HIs (fulfilling requirements, referred to as 'fitness'). It exclusively utilizes current guided wave data, eliminating the need for historical information. Ensemble learning techniques were also used to enhance HI quality while reducing deep learning randomness. The fitness equation is refined for dependable comparisons and practicality. The methodology is validated through investigations on T-single stiffener CFRP panels under compression-fatigue and dogbone CFRP specimens under tension-fatigue loadings, showing high performance of up to 93% and 81%, respectively, in prognostic criteria.

In this research, a generalized machine learning (ML) framework is proposed to estimate the fatigue life of epoxy polymers and additively manufactured AlSi10Mg alloy materials, leveraging their failure surface void characteristics. An extreme gradient boosting algorithm-based ML framework encompassing Synthetic Minority Over-sampling TEchnique (SMOTE), categorical data encoding, and external loop cross-validation is developed to evaluate the fatigue life across materials. The influence of different training strategies based on materials, input features, encoding method, and data standardization on the model performance is explored. Additionally, the importance of anti-data-leakage and anti-overfitting measures over the ML model performance is addressed. The result shows that the data-leakage-free, external loop cross-validated model can estimate the fatigue life of selective epoxy polymers and metal alloys with an average R² of 0.71 ± 0.06 using a mere 12 to 27 experimental data points per material category. Whereas the model trained with data-leakage and overfitting results in high R² of 0.9.

In a wide range of disciplines, such as aeronautical, automotive, and structural engineering, infrared thermography (IRT) has demonstrated promising performance for inspecting and monitoring structures. This chapter provides a survey of IRT as a nondestructive assessment method for composite materials subjected to impact loads that cause critical damage scenarios inside the structure. The chapter starts with a general overview of the principles of IRT and its historical evolution. Various methodologies of thermography inspection, such as optically, mechanically, and inductively stimulated thermography, are described for the evaluation of impact damage in composite structures. In the heating waveforms and data processing section, the relevant processing techniques and data analysis algorithms for each type of the above-mentioned thermography methodologies based on the applied waveform are presented. Case studies of composite specimens suffering from impact damage are provided at the end of the chapter to demonstrate the applicability and performance of IRT inspection.

The prognostic of the Remaining Useful Life (RUL) of composite structures remains a critical challenge as it involves understanding complex degradation behaviors while it is emerging for maintaining the safety and reliability of aerospace structures. As damage accumulation is the primary degradation indicator from the structural integrity point of view, a methodology that enables monitoring the damage mechanisms contributing to the structure's failure may facilitate a reliable and effective RUL prognosis. Therefore, in this study, an integrated methodology has been introduced by targeting the RUL and progressive delamination state via Deep Neural Network (DNN) trained with Guided wave-based damage indicators (GW-DIs). These GW-DIs are obtained via signal processing, Hilbert transform, and Continuous Wavelet Transform. This work uses GW-DIs to train and test the proposed model within two frameworks: one focusing on individual sample analysis to explore path dependency in RUL and delamination prognosis and another on an ensembled dataset to propose a generic model across varying stress scenarios. Results from the study indicate that proposed DNN frameworks are capable of encapsulating fast and slow degradation scenarios to evaluate the RUL prediction with associated delamination progress, which could contribute to ensuring the integrity and longevity of critical life-safe structures.

Health indicators are indices that act as intermediary links between raw SHM data and prognostic models. An efficient HI should satisfy prognostic requirements such as monotonicity, trendability, and prognosability in such a way that it can be effectively used as an input in a prognostic model for remaining useful life estimation. However, discovering or designing a suitable HI for composite structures is a challenging task due to the inherent complexity of the evolution of damage events in such materials. Previous research has shown that data-driven models are efficient for accomplishing this goal. Large labeled datasets, however, are normally required, and the SHM data can only be labeled, respecting prognostic requirements, after a series of nominally identical structures are tested to failure. In this paper, a semi-supervised learning approach based on implicitly imposing prognostic criteria is adopted to design a novel HI suitable. To this end, single-stiffener composite panels were subjected to compression-compression fatigue loading and monitored using acoustic emission (AE). The AE data after signal processing and feature extraction were fused using a multi-layer LSTM neural network with criteria-based hypothetical targets to generate an intelligent HI. The results confirm the performance of the proposed scenario according to the prognostic criteria.

The research shows the link between the strain transfer properties of distributed optical fiber sensors and their probability of damage detection, which is crucial for a successful implementation in real structural health monitoring applications.

Distributed Optical Fiber Sensors (DOFS) show several inherent benefits with respect to conventional strain-sensing technologies and represent a key technology for Structural Health Monitoring (SHM). Despite the solid motivation behind DOFS-based SHM systems, their implementation for real-time structural assessment is still unsatisfactory outside academia. One of the main reasons is the lack of rigorous methodologies for uncertainty quantification, which hinders the performance assessment of the monitoring system. The concept of Probability of Detection (POD) should function as the guiding light in this process, but precautions must be taken to apply this concept to SHM, as it has been originally developed for Non-Destructive Evaluation techniques. Although DOFS have been the object of numerous studies, a well-established methodology for their performance evaluation in terms of PODs is still missing. In the present work, the concept of Probability of Delamination Detection (POD²) is proposed for a DOFS network; Carbon Fiber-Reinforced Polymers (CFRP) Double-Cantilever Beam (DCB) specimens equipped with DOFS have been tested under static loading, and the strain patterns along with the relative observed delamination size have been collected to generate an adequate database for the POD analysis, suggesting a reference methodology to quantify the performance of DOFS for delamination detection.

In this paper we execute a complex test campaign to develop a novel methodology for the Remaining Useful Life (RUL) estimation of complex multi-stiffened composite aeronautical panels utilizing Machine Learning models trained with Structural Health Monitoring (SHM) data from hierarchically simpler elements, i.e., single-stiffened panels. Distributed Fiber Optical sensors (DFOS) are employed to monitor the panels’ behavior undergoing variable amplitude compression-compression fatigue after multiple impacts. A data processing methodology is first applied to the DFOS data, to both alleviate the effect of the variable loading conditions on the monitored strain and ease the computational burden. In this upscaling endeavor, an advanced strain-based Health Indicator (HI) based on Genetic algorithms, created and validated on the single-stiffened panel data, is utilized as the prognostic feature for the RUL estimations of the multi-stiffened panels. The HI displays favorable characteristics in terms of monotonicity and prognosability which are highly desirable for more accurate RUL estimations. For the prognostic task, standard machine learning models are trained using the historical degradation data of the single-stiffened panels and a similarity analysis is performed to enhance the accuracy when predicting the RUL of the multi-stiffened panels. Despite the increased structural complexity of the multi-stiffened panels, we demonstrate that the RUL is able to be predicted with reasonable accuracy. The present work paves the road for upscaling and applying prognostic methodologies to more complex structures beyond simple coupons or generic elements.