Surajit Roy

A novel physics-based temperature compensation model for structural health monitoring using ultrasonic guided waves

This article investigates the role of ambient temperature in causing changes to the structural wave propagation, as sensed by piezoelectric transducers, in a newer perspective. A novel approach is proposed to compensate the influence of temperature on piezo-sensor response using both analytical models and numerical simulations. Parametric studies using numerical simulations for plates with surface-mounted piezoelectric transducers establish linear functional relationship between change in sensor signals and specific combination of material properties, within certain temperature range. A numerical temperature compensation model is developed based on this functional relationship to reconstruct piezo-sensor signals at elevated temperatures. Matching pursuit–based signal analysis and reconstruction schemes are used in this study. Practical efficacy of the compensation model is tested for metallic structures with both simple and complex geometries. Model-based reconstruction of first wave packets in the sensor signals is found to match quite well with the experimental measurements. Performance of the proposed compensation model is also found to be at par with the existing state-of-art temperature compensation methods. A very limited set of baseline sensor data is required to estimate unknown model parameters, making this approach to be efficient and practically useful. The output of the compensation model is also used to obtain an accurate estimate of damage location in a structure under varying ambient temperature environments.

Recent advancements and vision toward stretchable bio-inspired networks for intelligent structures

Significant progress has recently been achieved in structural health monitoring, maturing the technology through quantification, validation, and verification to promote implementation and fielding of SHM. In addition, there is ongoing work seeking to detect damage precursors and to deploy structural health monitoring systems over large areas, moving the technology beyond hot-spot monitoring to global state sensing for full structural coverage. A large number of small sensors of multiple types are necessary in order to accomplish the goals of structural health monitoring, enabling increased sensing capabilities while reducing parasitic effects on host structures. Conventional sensors are large and heavy, adding to the weight of a structure and requiring physical accommodation without adding to and potentially degrading the strength of the overall structure. Increased numbers of sensors must also be deployed to span large areas while maintaining or increasing sensing resolution and capabilities. Traditionally, these sensors are assembled, wired, and installed individually, by hand, making mass deployment prohibitively time consuming and expensive. In order to overcome these limitations, the Structures and Composites Lab at Stanford University has worked to develop bio-inspired microfabricated stretchable sensor networks. Adopting the techniques of complementary metal-oxide semiconductor and microelectromechanical system fabrication, new methods are being developed to create integrated networks of large numbers of various micro-scale sensors, processors, switches, and all wiring in a single fabrication process. Then the networks are stretched to span areas orders of magnitude larger than the original fabrication area and deployed onto host structures. The small-scale components enable interlaminar installation in laminar composites or adhesive layers of built-up structures while simultaneously minimizing parasitic effects on the host structure. Additionally, data processing and interpretation capabilities could be embedded into the network before material integration to make the material truly multifunctional and intelligent once fully deployed. This article reviews the current accomplishments and future vision for these systems in the pursuit of state sensing and intelligent materials for self-diagnostics and health monitoring.

An integrated health management system for real-time impact monitoring and prediction of impact-induced damage on composite structures

Next generation technology of integrated health management systems for air-transportation structures will utilize SHM methods in combination with simulation techniques for the prediction of structural degradation induced by adverse events such as impacts. The contribution focuses on the development of an advanced real-time monitoring system for impact loads using passive sensing networks. Starting from the fundamental approach of real-time monitoring based on system identification models, problems of model order, signal conditioning and efficient model training will be addressed. Finally, the load monitoring system is interactively linked to a damage prediction module based on numerical failure analysis employing composite failure criteria. The utilization of appropriate database techniques allows a real-time prediction of impact induced damage after detection of any adverse impact event making information available on developing degradation at the earliest possible state.

Real-time prediction of impact-induced damage for composite structures based on failure analysis and efficient database methods

This contribution presents a novel strategy to achieve real-time prediction of non-penetrating impact-induced damage, especially delamination between plies of composite laminated structures. The proposed strategy aims to create an optimum-sized database of simulated damage information on a given laminated structure using numerical failure models and pattern recognition technique. A multi-stage data clustering algorithm is implemented to identify regions in the structural Finite Element (FE) model that have similar damage signatures. The generated database is linked to the real-time estimate of impact location and load-time history obtained from piezoelectric transducers based passive impact monitoring system. A composite stiffener panel is selected as a model problem and it is shown that the proposed strategy based on pattern recognition will result in large savings in computational cost for the database generation besides providing real-time damage diagnostic capabilities for in-service adverse impact events within certain confidence bounds.

Structural Health Monitoring of High Temperature Composites

High-temperature polymer-matrix composites (PMCs) are necessary and critical for the development of supersonic aircraft and orbital re-entry vehicles because of the need for light-weight design, high strength-to-weight ratios and high thermal stability in structures. Damage detection is a primary concern in composite structures because they are prone to multiple damage forms that can be hidden within the structure. Damage can include matrix cracking, fiber breakage, and delamination which can be caused by impacts, fatigue, or overloading. To overcome these shortfalls highly damage tolerant structures are employed to improve the safety of structures. Unfortunately this requires additional, potentially unnecessary, structural weight which is detrimental to aerospace structures. Acoustic ultrasound based structural health monitoring (SHM) has demonstrated the ability to overcome these problems by using arrays of Lead Zirconate Titanate piezoelectric transducers typically mounted on a flex circuit all of which is permanently affixed to, or embedded within, a structure [1] [2] [3] [4]. These transducers can excite and detect ultrasonic wave propagation in the structure and diagnostic algorithms, interpreting the signals, have been developed enabling real time inspection for damage. However, modern SHM systems are not capable of surviving the high temperatures experienced in the fabrication and service of High-temperature polymer matrix composites. In particular the Lead Zirconate Titanate piezoelectric elements typically depolarize and lose their functionality at around 200°C [5] [6]. Additionally, current SHM diagnostic algorithms are dependent on baseline data to compare signals to. These signals change with temperature and even just a few degree change can be detrimental to the system’s abilities. The current method for enabling functionality over a range of temperatures is to take numerous sets of baseline data at very high resolution across a range of temperatures. In order to adapt SHM for high temperature composites new piezoelectric materials must be developed capable of surviving elevated fabrication and operational temperatures. Small scale network components must be integrated to reduce detrimental effects of embedding SHM systems within the composite layup [7] [8] [9]. Additionally, methods for reducing the number of baseline data sets in the diagnostic algorithms must be developed. This paper presents development and testing of Bismuth Scandium Lead Titanate piezo ceramic transducers for high temperature SHM. These transducers are incorporated into a stretchable network system and mounted on a glass backing. Functionality is tested using a commercially available data acquisition system designed for SHM and intended for use with PZT transducers. Ongoing development of temperature compensation algorithms is also presented herein.© 2011 ASME

A Model-assisted Integrated Diagnostics for Structural Health Monitoring

This paper presents a novel model-assisted integrated diagnostics for structural health monitoring. Acousto-ultrasonic Lamb waves are propagated through a structure induced and sensed by an active piezoelectric sensor network. The damage is detected by comparing current sensor signals (with damage) with baseline signals (without damage). Diagnostic algorithm interprets changes in the signals to detect and localize damage. Typically the accuracy of damage localization depends on a priori knowledge of the velocity of Lamb waves. The estimation of Lamb wave velocity for complex structures is very challenging since analytical relations only exist for structures with simple geometries. Hence, a numerical tool based on spectral element method has been developed and employed to simulate acousto-ultrasonic wave propagation in structures. This tool generates an accurate velocity profile of Lamb waves propagating through a complex structure, which is used for offline training of the diagnostic algorithm. In order to achieve accurate diagnostics in varying temperature environments, the diagnostic algorithm has been integrated with a model to compensate for the effect of change in ambient temperature on sensor signals. Numerical tests are carried out to determine the performance of the integrated diagnostics for an aluminum stiffened panel with an open crack at different ambient temperatures. The results demonstrate that the proposed model-assisted integrated diagnostics has the capability of providing an accurate localization of damage in complex structures.

7.20 Structural Health Monitoring of Composites

The monitoring and subsequent management of structural health plays an essential role in composite materials and structures. Structural health monitoring (SHM) is an emerging technology which combines advanced sensor technologies with intelligent algorithms to interrogate the “health” condition of structures in real time or whenever necessary. Benefits of the SHM technology include improvement of reliability and safety, enhancement of performance and operation, potential for automated unsupervised monitoring, and reduction of life-cycle cost. Recent advancements in SHM technology have resulted in significant and promising techniques offering autonomous solutions for detecting and assessing the health condition of composite structures, providing a real-time early warning capability to mitigate damage, and reducing considerably downtime and maintenance costs. SHM technology can be applied to a wide range of systems within the aerospace, mechanical and civil engineering communities. This chapter examines the recent state-of-the-art SHM sensing, diagnostic and computational approaches for composite structures and the requirements for such applications.

Decision Making for Reference-Free Damage Detection

This paper concerns a reflection on a new decision boundary technique devoted for baseline-free damage detection purpose. Its scope focuses in studying analytically the impact of an unknown disturbance on the behavior of a monitored structure, through linear and nonlinear perturbation models. These perturbation models are introduced to interpret how an unknown disturbance away linearly and nonlinearly a monitored structure from its initial state, which is a healthy one, to its current state, which can be damaged or a healthy one that has undergone environmental/operational variations. To quantify the amount of that unknown disturbance effect, matrix perturbation theory is addressed to define two analytical bounds. The gap between them gives the decision regarding the monitored structure current state. The effectiveness of the proposed approach is demonstrated through an experimental setup of a test coupon aluminum plate, which has undergone varying loads and temperatures conditions, and crack damage cases. doi: 10.12783/SHM2015/367