In this work, a prognostics framework to predict the evolution of damage in fiber-reinforced composites materials under fatigue loads is proposed. The assessment of internal damage thresholds is a challenge for fatigue prognostics in composites due to inherent uncertainties, existence of multiple damage modes, and their complex interactions. Our framework, considers predicting the balance of release strain energies from competing damage modes to establish a reference threshold for prognostics. The approach is demonstrated on data collected from a run-to-failure tension-tension fatigue experiment measuring the evolution of fatigue damage in carbon-fiber-reinforced polymer (CFRP) cross-ply laminates. Results are presented for the prediction of expected degradation by micro-cracks for a given panel with the associated uncertainty estimates.

In this work, a framework for the estimation of the fatigue damage propagation in CFRP composites is proposed. Macro-scale phenomena such as stiffness and strength degradation are predicted by connecting micro-scale and macro-scale damage models in a Bayesian filtering framework that also allows incorporating uncertainties in the prediction. The approach is demonstrated on data collected from a run-to-failure tension-tension fatigue experiment measuring the evolution of fatigue damage in CRFP cross-ply laminates. Results are presented for the prediction of expected end of life for a given panel with the associated uncertainty estimates.

Composites offer unique advantages for aerospace structures and are increasingly being adopted into newer designs. However, it is also acknowledged that given current understanding of damage mechanisms in composites there is a significant risk with the extensive use of composites materials in aerospace applications. On one hand the uncertainty in damage evolution along lifetime is extremely large, and on the other hand there is a lack of knowledge about the mechanics of the onset, posterior growth, and interactions between several micro-scale damage modes. All these factors lead to the adoption of high safety margins in the design and costly inspection schedules along the service to mitigate the risks. Structural health monitoring for onboard damage diagnosis and prognosis of structural failures has the potential to reduce maintenance costs and improve the safety of the structure through a condition based maintenance scheduling. In this scheme the current damage state of a specific structural element is estimated and further used as the input for a prognostic algorithm that predicts the propagation of damage through time using updated models and based on some knowledge of the future load conditions. A novel damage prognostics framework for composites FRP under fatigue loadings is proposed in this work. The proposed methodology is grounded on physics-based models for evolution of damage at (1) micro-scale, i.e. micro-cracks and delamination, and (2) macro-scale such as stiffness reduction induced by micro-scale damage modes. Through stochastic embedding, these apriori deterministic models are converted to probabilistic models by introducing a modeling error term. This error term is controlled by a probability density function whose parameters are estimated in addition to the rest of "physical" parameters. The probabilistic damage models are then incorporated in a Bayesian filtering algorithm that sequentially updates both, a damage state variable and the set of model parameters, as fresh damage data become available along the fatigue cycling process. Next, these damage models are used to simulate fault propagation with this updated state information to generate a prognostic estimate of the remaining useful life of the structure in a probabilistic sense. The proposed methodology is demonstrated using experimental NDE damage data for micro-crack density, delamination area, and stiffness reduction from an extensive post-impact tension-tension fatigue test performed over several CFRP [0,90]4s laminates.

DYNAMIC STRAIN MAPPING AND REAL-TIME DAMAGE STATE ESTIMATION UNDER BIAXIAL RANDOM FATIGUE LOADING SUBHASISH MOHANTY*, ADITI CHATTOPADHYAY*, JOHN N. RAJADAS**, AND CLYDE COELHO* Abstract. Fatigue damage and its prediction is one of the foremost concerns of structural integrity research community. The current research in structural health monitoring (SHM) is to provide continuous (or on demand) information about the state of a structure. The SHM system can be based on either active or passive sensor measurements. Though the current research on ultrasonic wave propagation based active sensing approach has the potential to estimate very small damage, it has severe drawbacks in terms of low sensing radius and external power requirements. To alleviate these disadvantages passive sensing based SHM techniques can be used. Currently, few efforts have been made towards, time-series fatigue damage state estimation over the entire fatigue life (stage-I, II & III). A majority of the available literature on passive sensing SHM techniques demonstrates the clear trend in damage growth during the final failure regime (stage-III regime) or during when the damage is comparatively large enough. The present paper proposes a passive sensing technique that demonstrates a clear trend in damage growth almost over the entire stage-II and III damage growth regime. A strain gauge measurement based passive SHM frameworks that can estimate the time-series fatigue damage state under random loading is proposed. For this purpose, a Bayesian Gaussian process nonlinear dynamic model is developed to map the reference condition dynamic strain at a given instant of time. The predicted strains are compared with the actual sensor measurements to estimate the corresponding error signals. The error signals estimated at two different locations are correlated to estimate the corresponding fatigue damage state. The approach is demonstrated for an Al-2434 complex cruciform structure applied with biaxial random loading.

To improve the reliability and design of armor, it is imperative to understand the failure modes and the degradation rates of the materials used in armor. Despite the best efforts of manufacturers, some vulnerability of armor materials to ageing due to hydrolytic or oxidative environments is expected and may result in the degradation of material properties such as tensile strength. In this work, p-aramid yarns from two manufacturers were exposed to environmental conditions of various fixed temperature and humidity combinations. The maximum temperature and humidity condition was 70 °C and 76 % RH. Tensile tests were performed on specimens extracted at several different timepoints over the course of at least one year to determine the change in ultimate tensile strength and failure strain as a function of time, temperature, and humidity. These materials were found to be generally resistant to degradation at most conditions, showing changes of less than 10 % only at the highest temperature and humidity conditions. This data set contains failure load and failure strain values for three different aramid yarns, exposed to various conditions. It also includes Fourier Transform Infrared (FTIR) spectroscopy spectra for two of the aramids.

Model-based prognostics approaches capture system knowl- edge in the form of physics-based models of components that include how they fail. These methods consist of a damage estimation phase, in which the health state of a component is estimated, and a prediction phase, in which the health state is projected forward in time to determine end of life. However, the damage estimation problem is often multi-dimensional and computationally intensive. We propose a model decom- position approach adapted from the diagnosis community, called possible conflicts, in order to both improve the com- putational efficiency of damage estimation, and formulate a damage estimation approach that is inherently distributed. Local state estimates are combined into a global state esti- mate from which prediction is performed. Using a centrifugal pump as a case study, we perform a number of simulation- based experiments to demonstrate the approach.

The chapter describes the application of prognostic techniques to the domain of structural health and demonstrates the efficacy of the methods using fatigue data from a graphite-epoxy composite coupon. Prognostics denotes the in-situ assessment of the health of a component and the repeated estimation of remaining life, conditional on anticipated future usage. The methods shown here use a physics-based modeling approach whereby the behavior of the damaged components is encapsulated via mathematical equations that describe the characteristics of the components as it experiences increasing degrees of degradation. Mathematical rigorous techniques are used to extrapolate the remaining life to a failure threshold. Additionally, mathematical tools are used to calculate the uncertainty associated with making predictions. The information stemming from the predictions can be used in an operational context for go/no go decisions, quantify risk of ability to complete a (set of) mission or operation, and when to schedule maintenance.

Damage characterization through wave propagation and scattering is of considerable interest to many non-destructive evaluation techniques. For fiber-reinforced composites, complex waves can be generated during the tests due to the non-homogeneous and anisotropic nature of the material when compared to isotropic materials. Additional complexities are introduced due to the presence of the damage and thus results in difficulty to characterize these defects. The inability to detect damage in composite structures limits their use in practice. A major task of structural health monitoring is to identify and characterize the existing defects or defect evolution through the interactions between structural features and multidisciplinary physical phenomena. In a wave-based approach to addressing this problem, the presence of damage is characterized by the changes in the signature of the resultant wave that propagates through the structure. In order to measure and characterize the wave propagation, we use the response of the surface-mounted piezoelectric transducers as input to an advanced machine-learning based classifier known as a support vector machine.

Structural damage to ball grid array interconnects incurred during vibration testing has been monitored in the prefailure space using resistance spectroscopy-based state space vectors, rate of change of the state variable, and acceleration of the state variable. The technique is intended for condition monitoring in high reliability applications where the knowledge of impending failure is critical and the risks in terms of loss of functionality are too high to bear. Future state of the system has been estimated based on a second-order Kalman Filter model and a Bayesian Framework. The measured state variable has been related to the underlying interconnect damage in the form of inelastic strain energy density. Performance of the prognostic health management algorithm during the vibration test has been quantified using performance evaluation metrics. The method- ology has been demonstrated on leadfree area-array electronic assemblies subjected to vibration. Model predictions have been correlated with experimental data. The presented approach is applicable to functional systems where corner interconnects in area-array packages may be often redundant. Prognostic metrics including α − λ precision, β accuracy, and relative accuracy have been used to assess the performance of the damage proxies. The presented approach enables the estimation of residual life based on level of risk averseness.

The estimation of remaining useful life (RUL) of a faulty component is at the center of system prognostics and health management. It gives operators a potent tool in decision making by quantifying how much time is left until functionality is lost. This is especially true for aerospace systems, where unanticipated subsystem downtime may lead to catastrophic failures. RUL prediction needs to contend with multiple sources of error like modeling inconsistencies, system noise and degraded sensor fidelity. Bayesian theory of uncertainty management provides a way to contain these problems by integrating out the nuisance variables. We use the Relevance Vector Machine (RVM), for model development. RVM is a Bayesian treatment of the well known Support Vector Machine (SVM), a kernel-based regression/classification technique. This model is next used in a Particle Filter (PF) framework. Statistical estimates of the noise in the system and anticipated operational conditions are processed to provide estimates of RUL in the form of a probability density function (PDF). Validation of this approach on experimental data collected from Li-ion batteries is presented.