Michal Dziendzikowski

The Development of the Non-Parametric Classification Models for the Damage Monitoring on the Example of the ORLIK Aircraft Structure

This paper presents approach for the damage growth monitoring and early damage detection in the PZL-130 ORLIK TC II turbo propeller military trainer based on the array of the PZT sensors which will be deployed in the structure of the aircraft. Special attention will be paid to the preliminary results of the statistical models which provide an automated tool to infer about the damage presence and its size. In particular the effectiveness of the selected signal characteristics will be assessed using dimensional reduction methods (PCA) and the so called averaged damage indices will be described. Verification of the several classification models based on the emerged damage indices will be presented using cross validation techniques. The preliminary results of the data collected from the subcomponents tests with the model description, as well as approach for the SHM system design will be delivered. The verification of the models results will be presented on the example of the aerospace structures.

Damage Size Estimation of the Aircraft Structure with Use of Embedded Sensor Network Generating Elastic Waves

One of the approach to develop a system of continues, automated monitoring of the health of the structures is to use elastic waves excited in a given medium by piezoelectric transducers network. Elastic waves depending on their source and the geometry of the structure under consideration can propagate over significant distance. They are also sensitive to local structure discontinuities and deformations providing a tool to detect local damage of large aerospace structures. In the paper the issue of fatigue crack growth monitoring by means of elastic guided waves actuated by a sparse array of sensors will be presented. In particular we propose signal characteristics, robust enough to detect different kinds of damages: Barely Visible Impact Damages (BVIDs) in composite materials and fatigue cracks of metallic structures. The model description and the results of specimen tests verifying damage detection capabilities of the proposed signal characteristics are delivered in the paper. Some issues concerning the proposed damage indices and its application to damage detection and its monitoring are also discussed.

Energy Correlated Damage Indices in Fatigue Crack Extent Quantification

Signals received by piezoelectric transducers (PZT) network can be influenced by many factors. Apart from environmental conditions, whose variability should be compensated, significant difference in a signal can be also caused by relative geometry changes of a designed sensors node, e.g. the damage localization and its orientation with respect to sensors location in the node. In the adopted approach a set of damage indices (DIs), carrying marginal signal information content and correlated with the total energy received by a given sensor are proposed. These are sensitive to the two main modes of guided wave interaction with a fatigue crack, i.e. its transmission and reflection from a damage. Detailed description of DIs detection capabilities are delivered in the paper. Two dimensional reduction techniques: Principal Component Analysis and Fishers Linear Discriminant are compared. The results of the data collected from specimen fatigue test are used to compare several classification models based on the emerged effective damage indices.

Structural Health Monitoring of a Composite Panel Based on PZT Sensors and a Transfer Impedance Framework

One of the ideas for development of Structural Health Monitoring (SHM) systems is based on excitation of elastic waves by a network of PZT piezoelectric transducers integrated with the structure. In the paper, a variant of the so-called Transfer Impedance (TI) approach to SHM is followed. Signal characteristics, called the Damage Indices (DIs), were proposed for data presentation and analysis. The idea underlying the definition of DIs was to maintain most of the information carried by the voltage induced on PZT sensors by elastic waves. In particular, the DIs proposed in the paper should be sensitive to all types of damage which can influence the amplitude or the phase of the voltage induced on the sensor. Properties of the proposed DIs were investigated experimentally using a GFRP composite panel equipped with PZT networks attached to its surface and embedded into its internal structure. Repeatability and stability of DI indications under controlled conditions were verified in tests. Also, some performance indicators for surface-attached and structure-embedded sensors were obtained. The DIs’ behavior was dependent mostly on the presence of a simulated damage in the structure. Anisotropy of mechanical properties of the specimen, geometrical properties of PZT network as well as, to some extent, the technology of sensor integration with the structure were irrelevant for damage indication. This property enables the method to be used for damage detection and classification.

3D Reconstruction of Ultrasonic B-Scans for Nondestructive Testing of Composites

The paper presents an approach to 3D reconstruction of a sequence of ultrasonic B-Scans for the purpose of facilitating nondestructive testing of composites. The results of ultrasonic testing of a carbon fiber reinforced polymer specimen with barely visible impact damage was used for algorithm testing. 3D visualisation of damage based on image thresholding, contour extraction and volume rendering facilitates interpretation of ultrasonic data and can be useful in the assessment of a flaw size and location, including its depth. Accuracy of the 3D reconstruction of the internal damage of the tested specimen was verified on the basis of reference data acquired with the X-ray computed tomography. Owing to the low computational complexity of the proposed algorithm it could be applied during ultrasonic inspections of composite structures.

A method to compensate non-damage-related influences on Damage Indices used for pitch-catch scheme of piezoelectric transducer based Structural Health Monitoring

The risk of false calls of structural health monitoring systems is as much important for their application as their damage detection capabilities. Structural health monitoring based on guided waves propagation is particularly vulnerable to false calls. Signals acquired for piezoelectric transducer networks can be changed by many factors other than damage, for example, environmental factors or those related to the transducers’ aging or degradation of the transducers’ bonding with the monitored structure. A lot of studies were devoted to examine non-damage-related influences on structural health monitoring systems based on Lamb waves and to compensate the undesired effects. Most of compensation methods act on the level of the signal, that is, for a given factor influencing performance of piezoelectric transducers, signals are transformed to match the corresponding baselines. After such compensation procedure, the Damage Indices are calculated for the purpose of damage detection. In order to compensate the impact of all of non-damage-related factors, all of them need to be, at least, recognized. In this article, a different technique to compensate changes of Damage Indices values caused by factors other than damage is proposed. The method does not involve any operation on signals, but on the Damage Indices themselves. The factors causing Damage Indices’ changes neither have to be measured nor are even known. The capabilities of the method have been evaluated on the example of fatigue cracks detection in laboratory specimen tests and using results obtained during Full-Scale Fatigue Test of an aircraft.

Remote Monitoring of Fatigue Cracks Growth in the Aircraft Structure Based on Active Piezosensor Network during the Full Scale Fatigue Test

One of the major issues from a structural integrity point of view of the aircraft structure is an appropriate health monitoring technology delivery for the damage tolerant philosophy. This paper presents a development of a system for fatigue crack growth monitoring and early damage detection in the PZL 130 ORLIK TC II turbo-prop military trainer aft structure. The maintenance system of the aircraft shifts from the safe-life to the hard-time. The aircraft started Full Scale Fatigue Test (FSFT) which will continue up to 2013. In the article a built block approach for the system design, signal modeling, sensing and signal processing as well as damage detection is presented. Taking into the consideration a previous experience of AGH as well as AFIT, a network of PZT transducers was deployed in the aircraft structure hot-spots. The system components are: remote monitoring unit, signal analysis, graphical user interface, sensor self-diagnostic tools, and data classification model. Description of damage detection capabilities are delivered in the paper. In particular some issues concerning the proposed damage indices and its application to crack growth estimation models are discussed. Fishers Linear Discriminant is used as a method to obtain effective crack growth predictors and one of the self-diagnostic tools used in the system. The results of the data collected from specimen fatigue tests are delivered and cross-validation technique is used to evaluate a classification model based on the damage indices derived.

An Approach to Damage Detection in the Aircraft Structure with the Use of Integrated Sensors – The Symost Project

Abstract This paper presents an approach to damage growth monitoring and early damage detection in the structure of PZL - 130 ORLIK TC II turbo-prop military trainer aft using the statistical models elaborated by the Polish Air Force Institute of Technology (AFIT) and the network of the sensors attached to the structure. Drawing on the previous experiences of the AFIT and AGH in structural health monitoring, the present research will deploy an array of the PZT sensors in the structure of the PZL -130 Orlik TC II aircraft. The aircraft has just started Full Scale Fatigue Test (FSFT) that will continue up to 2013. The FSFT of the structure is necessary as a consequence of the structure modification and the change of the maintenance system - the transition to Condition Based Maintenance. In this paper, a novel approach to the monitoring of the aircraft hot-spots will be presented. Special attention will be paid to the preliminary results of the statistical models that provide an automated tool to infer about the presence of damage and its size. In particular, the effectiveness of the selected signal characteristics will be assessed using dimensional reduction methods (PCA) and the so-called averaged damage indices will be delivered. Moreover, the results of the signal classification based on the neural network will be presented alongside the numerical model of the wave propagation. The work contains selected information about the project scope and the results achieved at the preliminary stage of the project

Transfer Impedance Approach to Damage Detection and Localization Based on RAPID Imaging Algorithm

One of the ideas for structural health monitoring (SHM) systems built is based on analysis of small displacements propagation excited in the element by a network of PZT piezoelectric actuators. Structural damages can result in observable changes of the signal generated by the network sensors, due to elastic wave interaction with damage. The electromechanical impedance (EMI) method approach to SHM, harmonic excitation of PZT is used, thus steady elastic waves are excited in the structure. The signal can be gather in the pulse – echo scheme, i.e. when single transducer is used both as a actuator and the receiver of waves, as well in the pitch – catch scheme, when a pair of transducers, the generator and the sensor, are used. For the latter EMI approach, the term Transfer Impedance Approach is sometimes used. In the paper, an approach for damage detection and localization with use of network of PZT sensors excited with harmonic signals in broad frequency spectrum is presented. In particular, some signal characteristics – called Damage Indices (DI’s) used for structure assessment are proposed and their properties are discussed. The DI’s have the property to assess the location a damage with respect to a single sensing path, formed by a pair of transducers, i.e. the generator and the receiver. Finally, the RAPID algorithm for damage localization with use of the information from all of the network sensing paths is applied.

Predictive Models for Transient Loads of Vertical Stabilizer of an Aircraft Using Canonical Correlation Analysis

The knowledge about loads of the structure occurring during aircraft operation, is one of the fundamental element of the so called damage tolerance approach to aircraft design. In the optimal case, such information could be available from a network of sensors, e.g. strain gauges, FBGs, deployed in the aircraft structure and measuring its local stress. However, systems of operational loads monitoring (OLM) are still not widely applied. Instead, what is available, is a set of flight parameters, which by the laws of inertia and aerodynamics should determine dominant part of loads, acting on a given element. In this paper, canonical correlation analysis (CCA) will be discussed as an useful method for selection of flight parameters proper for prediction of aircraft loads.