Nanoparticle- (NP-) doped optical fibres show the potential to increase the signal-to-noise ratio and thus the sensitivity of optical fibre strain detection for structural health monitoring. In this paper, our previous experimental/simulation study is extended to a design study for strain monitoring. 100 nm spherical gold NPs were randomly seeded in the optical fibre core to increase the intensity of backscattered light. Backscattered light spectra were obtained in different wavelength ranges around the infrared C-band and for different gauge lengths. Spectral shift values were obtained by cross-correlation of the spectra before and after strain change. The results showed that the strain accuracy has a positive correlation with the relative spectral sensitivity and that the strain precision decreases with increasing noise. Based on the simulated results, a formula for the sensitivity of the NP-doped optical fibre sensor was obtained using an aerospace case study to provide realistic strain values. An improved method is proposed to increase the accuracy of strain detection based on increasing the relative spectral sensitivity, and the results showed that the error was reduced by about 50%, but at the expense of a reduced strain measurement range and more sensitivity to noise. These results contribute to the better application of NP-doped optical fibres for strain monitoring.

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 present work aims at understanding the stochastic matrix crack evolution in CFRP cross-ply laminates under tension–tension fatigue loading. An experimental campaign was carried out on twenty-three specimens at different stress levels, while two optical techniques were used for the in-situ monitoring of the accumulation of transverse matrix cracks. The results showed a significant scatter in crack evolution among specimens. This stochastic behaviour was further investigated using image analysis and numerical modelling. It was found that transverse matrix cracks can be classified into the independent and dependent cracks based on a critical crack spacing. Furthermore, the severity of interaction among cracks was quantified by introducing a dependent crack ratio. Finally, a strength-based probabilistic model was proposed to describe the scattering regime of the crack evolution. The agreement between model and test results indicates that local strength variations of 90 plies are the dominant scattering source governing the initial fatigue resistance to cracking and determining the accumulation of transverse matrix cracks among specimens. These results may provide a new insight into the stochastic nature of matrix cracking in composite laminates and aid in the design of fatigue resistance properties.

Embedding fiber optic sensors (FOSs) within parts for strain measurement is attracting widespread interest due to its great potential in the field of structural health monitoring (SHM). This work proposes a novel method of embedding FOSs using capillaries within solid structures and investigates fiber positions and orientation uncertainties within capillaries of different sizes and their influences on strain measurement accuracies. To investigate how the fiber positions and orientation variations influence strain measurement accuracy, both analytical and numerical models are utilized to predict strain distributions along embedded fibers at different positions and with different orientations within the specimen. To verify the predictions, a group of specimens made of Aluminum 6082 was prepared, and the specimens in each group had capillaries of 2 mm, 4 mm, and 6 mm diameters, respectively. Fibers were embedded within each specimen using the capillaries. Four-point bending static tests were conducted for each specimen with embedded FOSs, performing in situ strain measurement. Subsequently, the specimens were partitioned into several pieces, and the cross sections were observed to know the real positions of the embedded fiber. Finally, the strain predictions at the real locations of the fiber were compared with the measured strain from the embedded FOSs. The predicted strain distributions as a function of the fiber positions alone and as a function of both the fiber positions and orientations were compared to assess the influence of fiber orientation change. The results from a combination of analytical, numerical, and experimental techniques suggest that the fiber position from the capillary center is the main factor that can influence strain measurement accuracies of embedded FOSs, and potential fiber misalignments within the capillary had a negligible influence. The fiber position-induced measured error increases from 10.5% to 18.5% as the capillary diameter increases from 2 mm to 6 mm. A 2 mm capillary diameter is able to lead to the lowest measurement error in this study and maintains ease of embedding. In addition, it is found that the measured strain always lies within a strain window defined by the strain distribution along capillary boundaries when there are no cracks. This can be further studied for crack detection.

Under in-plane compressive load conditions, the growth of a delamination initially induced by an impact can be followed by a fast growth after a threshold level, which leads to a catastrophic failure in composite structures. To avoid reaching this critical level, it is essential to uncover the delamination size and growth pattern in real time. Ultrasonic Guided Waves (UGW) have a strong capability to interrogate and monitor the structure in real-time and thus track the growth of damage, which may occur during the flight cycles. Although various types of damage affect the monitored UGW signals, it is challenging to determine from the UGW signals what types of damage and at what rate of growth are occurring within the structure. UGW signals can be acquired at defined intervals and then analysed to possibly detect different types of damages, such as delamination, and to quantify the rate of damage growth over fatigue cycles. However, correlating the UGW-based Damage Indicators (DIs) with the specific type of damage, such as delamination, and damage growth is a challenging task as the relation between these DIs and the actual damage state is very complex. Therefore, in this study, a supervised Deep Neural Network-based (DNN) prediction model is proposed aiming to diagnose the delamination size of the composite structure by correlating the UGW-based DIs with the quantified time-varying delamination size. UGW data is collected through a network of permanently installed piezoelectric transducers (PZTs). The delamination size is obtained through ultrasonic C-Scan technique at defined cycles. DIs are extracted in time, frequency, and time-frequency domains and used as the input for the DNN-based regression model. Each sensor-actuator path is considered as an independent set of indicators, which are separated for training, validation, and testing purposes. The effect of the different paths on the delamination size prediction is presented along with the model performance on measured delamination growth in woven type composite sample.

A health indicator (HI) is a valuable index demonstrating the health level of an engineering system or structure, which is a direct intermediate connection between raw signals collected by structural health monitoring (SHM) methods and prognostic models for remaining useful life estimation. An appropriate HI should conform to prognostic criteria, i.e., monotonicity, trendability, and prognosability, that are commonly utilized to measure the HI's quality. However, constructing such a HI is challenging, particularly for composite structures due to their vulnerability to complex damage scenarios. Data-driven models and deep learning are powerful mathematical tools that can be employed to achieve this purpose. Yet the availability of a large dataset with labels plays a crucial role in these fields, and the data collected by SHM methods can only be labeled after the structure fails. In this respect, semi-supervised learning can incorporate unlabeled data monitored from structures that have not yet failed. In the present work, a semi-supervised deep neural network is proposed to construct HI by SHM data fusion. For the first time, the prognostic criteria are used as targets of the network rather than employing them only as a measurement tool of HI's quality. In this regard, the acoustic emission method was used to monitor composite panels during fatigue loading, and extracted features were used to construct an intelligent HI. Finally, the proposed roadmap is evaluated by the holdout method, which shows a 77.3% improvement in the HI's quality, and the leave-one-out cross-validation method, which indicates the generalized model has at least an 81.77% score on the prognostic criteria. This study demonstrates that even when the true HI labels are unknown but the qualified HI pattern (according to the prognostic criteria) can be recognized, a model can still be built that provides HIs aligning with the desired degradation behavior.

In this paper, the effect of interface properties on the compressive failure behavior of 3D woven composites (3DWC) is investigated by incorporating a micromechanics-based multiscale damage model (MMDM). The correlation between the mesoscopic stress of yarns and microscopic stress of constituents is established by defining a stress amplification factor. With the microscopic stresses, the fiber breakage and matrix failure can be separately evaluated at the microscale, without assuming the yarns as transversely isotropic homogeneous materials. Especially, the interfacial debonding between yarns and matrix is also a dominant damage mode within 3DWC. Given that there is still a lack of studies on the influence of interfacial properties on the compressive failure behavior of 3DWC, it is meaningful to perform numerical parametric studies to reveal how the interface properties contribute to the damage mechanisms of 3DWC under compressions. The predicted results indicate that with the increase of interface strengths and fracture toughness, the compressive resistance of 3DWC can be significantly improved, resulting in higher strength and failure strain. Additionally, the studied 3DWCs with weak, medium and strong interfaces exhibit different damage development processes.

A health indicator (HI) serves as an intermediary link between structural health monitoring (SHM) data and prognostic models, and an efficient HI should meet prognostic criteria, i.e., monotonicity, trendability, and prognosability. However, designing a proper HI for composite structures is a challenging task due to the complex damage accumulation process during operational conditions. Additionally, designing a HI that is independent of historical SHM data (i.e., from the healthy state until the current state) is even more challenging as HI and remaining useful life prediction are time-dependent phenomena. A reliable SHM technique is required to extract informative time-independent data, and a powerful model is necessary to construct a proper HI from that data. The lamb wave (LW) technique is a useful SHM method that can extract such time-independent data. However, translating the LW data at each time step to the appropriate HI value is a challenge. AI—deep learning in this case—offers significant mathematical potential to meet this difficulty. A semi-supervised learning approach is developed to train the model using the simulated ideal HIs as the targets. The model uses the current LW data, without prior or subsequent data, to output the current HI value. Prognostic criteria are then calculated using the entire HI trajectory until the end-of-life. To validate the proposed approach, aging experiments from NASA’s prognostics data repository are used, which include composite specimens subjected to a tension-tension fatigue loading and monitored using the LW technique. The LW data is first processed using the Hilbert transform, and then, upper and lower signal envelopes in two states – baseline and current time – are used to feed the deep learning model. The results confirm the effectiveness of the proposed approach according to the prognostic criteria. The effect of different triggering frequencies of the LW system on the results is also discussed in terms of the prognostic criteria.

To move towards a condition-based maintenance practice for aircraft structures, design of reliable health management methodologies is required. Development of diagnostic methodologies is commonly realised on simplified sample structures with assumptions that methodologies can be adapted for application to realistic aircraft structures under in-service conditions. Yet such actual applications are not conducted. In this work, we study the development of diagnostic methodologies to training structures and their application to dissimilar testing structures. A heterogeneous population is considered, consisting of single-stiffener composite panels for methodology development and training and a multi-stiffener composite panel for application and testing. Characteristics as its composite material, lay-up, and temperature condition are constant while topologies and applied loads differ between the dissimilar structures. Damage in the structural panels is monitored on multiple diagnostic levels using a variety of structural health monitoring (SHM) techniques, including acoustic emission and distributed strain sensing. Specifically, we develop diagnostic methods for localising and monitoring disbond growth after impact using strain data collected during fatigue testing of multiple single-stiffener panels and apply these for disbond monitoring in an upscaled version of a multi-stiffener panel. In this manner, this study aids in the maturement and application of SHM methodologies to realistic aircraft structures.

Ply-block size and stress level effect on accumulation of transverse cracks and delamination are investigated during early fatigue life of CFRP laminates. Tension-tension fatigue tests under different stress levels were performed for two cross-ply configurations. Edge observation with digital cameras, digital image correlation and acoustic emission were employed for in-situ damage monitoring. Transverse cracks were dominant for [0/90₂]_s laminates with almost non-existent delamination, while different interactive levels between both damage mechanisms occurred for [0₂/90₄]_s laminates. Poisson's ratio identifies whether early fatigue damage is dominant by transverse cracks or involves delamination. Cumulative AE energy is a helpful indicator of crack density.

In this paper, a reliable progressive fatigue damage model (PFDM) for predicting the fatigue life of composite laminates is proposed by combining the normalized fatigue life model, nonlinear residual degradation models and fatigue-improved Puck criterion. To balance the accuracy of life predictions and computational efficiency, an adaptive cyclic jump algorithm is developed and implemented within the PFDM. The sensitivity of life prediction to cyclic jump parameter has been greatly reduced by correlating the cyclic jump with the increment time and viscous coefficient. Therefore, the cyclic jump parameter can be arbitrarily selected within a relatively large range to obtain convergent results. When incorporating the adaptive cyclic jump algorithm, there is no need to define a standard for determining the material failure in numerical calculations, which effectively eliminates an artificially induced uncertainty in life predictions. Two sets of experiments are conducted to validate the proposed PFDM. The numerical predictions including static failure strength and fatigue life correlate reasonably well with the available experimental data.

For pick-and-place processes to become widely implemented in industry a consistent and acceptable product quality needs to be achieved. One important quality criterion is the fiber angle deviations in the reinforcement. Handling a reinforcement will subject it to forces due to e.g. gravity and accelerations. These forces can result in in-plane shear and subsequently in fiber angle deviations. The current work looks at predicting and preventing in-plane shear induced fiber angle deviations by studying the positioning of pick-up points. In the state of the art the positioning of individual pick-up points is typically either not discussed or is based on the mould where the fabric is to be draped on – not on the effect of the handling on the fabric. The relationship between the positioning of the pick-up points and the behavior of the fabric should however also be considered. Finite element simulations validated through experimental work will be used to study the influence of pick-up point location on the in-plane shear strain for a bi-axial Non Crimp Fabric [NCF]. In [1] tolerances have been set for the fiber angle deviations, additionally the relationship between in-plane shear strain and fiber angle deviations has been demonstrated for the specific NCF. These results are used in the present work to evaluate the results from the simulations. The current work will demonstrate that it is possible to control the in-plane shear induced fiber angle deviations by varying the position and number of the pick-up points. Additionally, It will show whether simulations for the positioning of pick-up points on large reinforcements can be simplified by looking at one instance of a repeating pattern. The paper will provide a framework for the determination of positioning of pick-up points while predicting and preventing in-plane shear induced fiber angle deviations.

Continuous ultrasonic welding is a promising high-speed and energy-efficient joining technique for thermoplastic composite structures. However, in the current state-of-the-art research on the topic numerous deconsolidation voids could be identified at the welding interface, which results in a strength knock-down. The aim of this study is, therefore, to improve the quality of continuous ultrasonically welded joints by adding a consolidation shoe to the welding setup. To determine the required consolidation pressure, the size of the shoe, and its distance from the sonotrode a stepwise approach was followed based on the static ultrasonic welding process. The closest consolidation distance, best representing the static welding conditions, did not improve the weld quality as significant porosity was still found in the weld line and in the adherends. However, for the furthest consolidation distance high-quality continuous welds were obtained with almost no porosity and a high strength.

Recently, companies all over the world have been focusing on the improvement of autonomous health management systems in order to enhance performance and reduce downtime costs. To achieve this, the remaining useful life predictions have been given remarkable attention. These predictions depend on the proper designing process and the quality of health indicators (HI) generated from structural health monitoring sensors based on prior established multiple prognostic evaluation criteria. Constructing such HIs from noisy sensory data demands powerful models that enable the automatic selection and fusion of features taken from those relevant measurements. Deep learning models are promising to autonomously extract features in scenarios with a huge volume of data without requiring considerable domain expertise. Nonetheless, the features established by artificial neural networks are complicated to comprehend and cannot be regarded as physical system characteristics. In this regard, the goal of this paper is to extend a new model; an interpretable artificial neural network that enables the automatic selection and fusion of features to construct the most appropriate HIs with remarkably fewer parameters. This model consists of additive and multiplicative layers that provide a feature fusion that better reflects the system’s physical properties. Additionally, the weights are discretized in two ways: a) using a ternary form with values {-1, 0, 1}, and b) relaxing the aforementioned ternary form by rounding the weights at the first decimal point in the range of [-1, 1]. Both discretization techniques have the ability to softly control the number of parameters that should be ignored. This trick guarantees interpretability for the neural network by extracting simple yet powerful equations representing the constructed HIs. Finally, the model’s performance is evaluated and compared with other approaches using a practical case study. The results show that the proposed approach's designed HIs are both interpretable and of high quality according to the criteria of the HI's evaluation.

With the increased use of composites in aircraft, many new successful contributions to the advancement of the structural health monitoring (SHM) field for composite aerospace structures have been achieved. Yet its application is still not often seen in operational conditions in the aircraft industry, mostly due to a gap between research focus and application, which constraints the shift towards improved aircraft maintenance strategies such as condition-based maintenance (CBM). In this work, we identify and highlight two key facets involved in the maturing of the SHM field for composite aircraft structures: (1) the aircraft maintenance engineer who requires a holistic damage assessment for the aircraft’s structural health management, and (2) the upscaling of the SHM application to realistic composite aircraft structures under in-service conditions. Multi-sensor data fusion concepts can aid in addressing these aspects and we formulate its benefits, opportunities, and challenges. Additionally, for demonstration purposes, we show a conceptual design study for a fusion-based SHM system for multi-level damage monitoring of a representative composite aircraft wing structure. In this manner, we present how multi-sensor data fusion concepts can be of benefit to the community in advancing the field of SHM for composite aircraft structures towards an operational CBM application in the aircraft industry.

Prediction models for fatigue in engineering applications are developed within a fatigue analysis framework, deliberately selected in some cases, but mostly chosen without substantiation. The proposition of this paper is that selecting the most appropriate framework can only be done with the knowledge and a complete overview of existing frameworks and their systematic categorization. In particular for composite materials, where due to coexistence of different mechanisms and their complex interaction under fatigue loading, only a unified approach can characterize the complete fatigue phenomenon. To that aid, this paper provides a complete overview of existing fatigue analysis frameworks for various materials along with such systematic categorization. Each analysis framework is based on a specific methodology that evolved over time. Hence, this overview is provided following the time stamp evolution of each methodology within different analysis frameworks. With such an overview, one can conclude that for fatigue analysis of composite materials, the theory of the thermodynamics of the irreversible processes and continuum damage mechanics framework provides the required unified approach. Additionally, this paper demonstrates that many material classes, like metals and composites, can be analysed using a common framework. This common framework has similarity up to a certain level, and at the detailed level, it differs by addressing the difference in material class-specific mechanisms.

In this paper, with the final aim of shape sensing for a morphing aircraft wing section, a developed multimodal shape sensing system is analysed. We utilise the method of interrogating a morphing wing section based on the principles of both hybrid interferometry and Fibre Bragg Grating (FBG) spectral sensing described in our previous work. The focus of this work is to assess the measurement performance and analyse the errors in the shape sensing system. This includes an estimation of the bending and torsional deformations of an aluminium mock-up section due to static loading that imitates the behaviour of a morphing wing trailing edge. The analysis involves using a detailed calibration procedure and a multimodal sensing algorithm to measure the deflection and shape. The method described In this paper, uses a standard single core optical fibre and two grating pairs on both the top and bottom surfaces of the morphing section. A study on the fibre placement and recommendations for efficient monitoring is also included. The analysis yielded a maximum deflection sensing error of 0.7 mm for a 347 × 350 mm wing section.

For pick-and-place processes to become widely implemented in industry a consistent and acceptable product quality needs to be achieved. In the state of the art it is assumed that reinforcements will be in perfect condition at the start of forming or draping. In reality the handling process can already result in undesired deformations. The current work will look at fiber angle deviations that occur during this process due to in-plane shear. It is shown that bounds can be set for these fiber angle deviations based on experimental work. Periodic representative volume element homogenization is used to obtain homogenized material properties for a bi-axial non-crimp fabric with a specific construction. With these material properties the in-plane shear strain, and thus the fiber angle deviations, can be predicted. The presented methodology and results obtained using it can be a basis in the design process for automated handling of reinforcements and for in-situ quality control of the pick-and-place process.

Nanoparticle (NP) doping of optical fibres can be used to increase the intensity of the backscattered light used for distributed strain sensing and has shown the advantages of high precision strain detection and multiplex sensing experimentally. However, the backscatter spectral characteristics of NP-doped optical fibres have not been described even though they are quite different from the spectra from fibre Bragg gratings (FBGs) or commercial single mode fibres. In this paper, gold NPs, used as the contrast agent in the optical fibre to increase the intensity of the backscattered light, were investigated from the aspect of their spectra. A single scattering model with Mie theory and an effective refractive index (RI) model were used to evaluate the backscattered light spectra and the Monte Carlo Method was used for seeding NPs. The results showed that the strain responsivity of gold-NP doped fibres with low volume ratio doping (single scattering restriction) are close to FBGs and commercial fibres. High volume ratios of gold NP doping increase the imaginary part of the RI of the optical fibre, which has a significant influence on the spectra in the wavenumber domain. These theoretical insights may promote the future engineering design of NP-doped fibre sensors.

The plasmon resonance spectral peak of a gold spherical nanoparticle (NP) will shift when the NP shape is changed from sphere to spheroid. This may be used as a novel strain detection method with gold NPs embedded in a medium of different refractive index (RI). Applying a strain to the external medium will cause a change in the shape of the NP from spherical to spheroidal. In our previous work, it was found that when the RI change of the medium surrounding the NPs is close to zero, the shape change induced plasmon resonance spectral peak shift will become important. In order to obtain only the wavelength shift values caused by the shape change of the NPs, the RI of medium surrounding the gold NPs is set at a constant of 1.45 and the RI of the gold NP is assumed unchanged. The T-matrix method is used to calculate the scattered light and light extinction by the NP morphing. The diameters of the gold NPs are set from 100 nm to 400 nm, with the size interval at 10 nm, to cover a wide size range for typical sizes of gold spherical NPs. The spectra of the light scattering and light extinction were calculated on the Delft University high performance computing cluster. The results show that the plasmon resonance spectral peak shift is related to the size of the NPs. Larger sizes of gold NPs have larger peak shift values, but there is an inflection point around 200 nm and the bandwidth of the resonance peak becomes larger which will cause a difficulty in precisely locating the peak.