Higher operational loads, greater complexity of design and longer lifetime periods imposed in civil, mechanical and aerospace structures, make it increasingly important to monitor the health of these structures. A wide variety of highly effective non-destructive methods, such as acoustic or ultrasonic methods, magnet field methods, penetrating liquids, eddy-current methods or thermal field methods, and so on, are currently available for the detection of defects. Unfortunately, they are all localized techniques, implying long and expensive inspection time; often, structural components are not inspected just because of their inaccessibility and damage can propagate to critical levels between the inspection intervals. The drawbacks of current inspection techniques have led engineers to investigate new methods for continuous monitoring and global condition assessment of structures. That is the case for methods based on vibration responses that allow one to obtain meaningful time and/or frequency domain data and calculate changes in the structural and modal properties, such as resonance frequencies, modal damping and mode shapes, FRF based methods, and use them with the objective of developing reliable techniques to detect, locate and quantify damage. The fundamental idea of the vibration methods for damage detection is that modal parameters (natural frequencies, mode shapes, and modal damping) are functions of the physical properties of the structure (mass, damping, and stiffness). Therefore, changes in the physical properties will cause detectable changes in the modal properties. So, this process involves the observation of a structure/system over time using periodical measurements. In other words, most vibration-based damage detection methods can be considered as some form of the pattern recognition problem as they look for the discrimination between two or more signal categories, e.g. before and after a structure is damaged or differences in the damage levels or locations. The work here presented addresses the subjects of damage detection, localization and quantification in structures. The reported examples show the implementation and comparison of a number of various damage detection and localization methods based on vibration responses changes. The objective of such a study is to ascertain the possibility of using various damage detection and localization methods with and without the need for modal identification. In recent years, the authors have developed some simplemethods and tools based upon the use of mode shapes and frequency response functions (FRFs), which seem promising and have given good results in some practical applications. The advantage of using such methods, which is based on mode shapes, is that only measured mode shapes are required in damage identification, without having to know the complete stiffness and mass matrices of the structure. To use of directly measured FRFs data, which provide an abundance of information, is further beneficial as the execution of experimental modal analysis is not needed, thus greatly reducing human induced errors. Here, all methods are first tested on data of a simple steel beam structure to assess their feasibility and performance. Then all proposed methods are applied to a more complicated structure, a typical aircraft stiffened panel, for extension and validation purposes. All developed methodologies are verified by numerical simulations and laboratory testing for both structure. As defects, notch type damage of different severity are investigated for the beam structure. For the panel structure four different types of structural change are studied. i.e. remove some screw for disconnecting the stiffener from the panel and create a saw cut along the all the depth of the hight of second stiffener. FRFs of panel in each state are recorded by two measurement systems (LMS Hammer Testing and PSV-Laser Scanner Vibrometer) and for each scenarios the related mode shapes are extracted by PolyMAX algorithm of LMS Testlab software. In addition, numerical simulated examples, after correlation with the real experimental one, of both mentioned structures will be used, as well as an experimental test cases, where damage is inflicted in a free-free condition. To simulate noise disturbances experienced during the experimental testing, for the numerical simulations of example beam, measurement data generated by MSC/Nastran are polluted with 3% of random noise level. The study shows the potential of the proposed methods for simple and rapid detection, accurate and reliable localization and monitoring of damage in structures. The reported examples also show that some proposed methods (i.e. PCA based, Transmissibility based) are highly capable for damage detection and localization and structural health monitoring even in the noise polluted data.
Carichi operativi elevati, maggiore complessità di progettazione e periodi di lavoro di maggiore durata rendono sempre più importante il monitoraggio dello stato di strutture meccaniche e aerospaziali civili. Un'ampia varietà di metodi non distruttivi altamente efficaci, quali metodi acustici o ad ultrasuoni, metodi che sfruttano campi magnetici, liquidi penetranti, metodi a correnti parassite e così via, sono attualmente disponibili per la rilevazione di difetti strutturali. Purtroppo, queste sono tutte tecniche localizzate, implicando lunghi e costosi tempi di ispezione. Spesso, i componenti strutturali non vengono ispezionati solo a causa della loro inaccessibilità ed in questo modo eventuali danni possono propagarsi fino a livelli critici tra gli intervalli di ispezione. Gli svantaggi delle tecniche di ispezione in utilizzo oggigiorno hanno portato allo studio di nuovi metodi per il monitoraggio continuo della condizione globale delle strutture. Questo è il caso dei metodi basati sulle vibrazioni che permettono di ottenere dati nel dominio del tempo e della frequenza che vengono utilizzati per calcolare variazioni delle proprietà strutturali e modali, come le frequenze di risonanza, smorzamento e le forme modali, FRF, e li usa con l'obiettivo di sviluppare tecniche affidabili per rilevare, localizzare e quantificare eventuali danni. L'idea fondamentale dei metodi basati sulle vibrazioni per il rilevamento danni è che i parametri modali (frequenze naturali, forme modali e smorzamento modale) siano funzioni delle proprietà fisiche della struttura (massa, smorzamento, e rigidezza). Pertanto, i cambiamenti nelle proprietà fisiche causeranno sostanziali cambiamenti nelle proprietà modali. Quindi, questo processo comporta l'osservazione di un sistema nel tempo con misurazioni periodiche. In altre parole, più metodi di rilevamento dei danni basati sulle vibrazioni possono essere considerate come una forma di problema di identificazione di modello in viene analizzato il discriminante tra due o più categorie di segnale, per esempio prima e dopo che la struttura sia stata danneggiata o eventuali differenze nei livelli di danno. Il lavoro qui presentato indirizza il problema del rilevamento dei danni strutturali, la loro localizzazione e quantificazione. Gli esempi riportati mostrano l’implementazione e confronto di diversi metodi di rilevamento e localizzazione del danno basati sulle variazioni delle risposte di vibrazione. L'obiettivo di questo studio è quello di verificare la possibilità di utilizzare diversi metodi di rilevamento e localizzazione del danno con e senza la necessità dell'identificazione modale. Negli ultimi anni, gli autori hanno sviluppato alcuni semplici metodi e strumenti basati sull'impiego di forme modali e funzioni di risposta in frequenza (FRF), le quali sembrano promettenti e hanno dato buoni risultati in alcune applicazioni pratiche. Il vantaggio di utilizzare tali metodi, che si basano su forme modali, è che quest’ultime, quando misurate, sono solo necessarie nell’identificazione del danno, senza dover conoscere le matrici di rigidezza e massa della struttura. L’utilizzo diretto della FRF, che fornisce un numero elevato di informazioni, è di ulteriore utilità quando non è possibile l'esecuzione di un’ analisi modale sperimentale, quindi riducendo notevolmente gli errori umani indotti. In questo lavoro, tutti i metodi sono prima stati provati su dati di una trave di acciaio a semplice per valutare la fattibilità del metodo e le sue prestazioni. I metodi proposti sono poi provati per una struttura più complicata, un tipico pannello aeronautico con irrigidimenti, per scopi di estensione e di validazione. Tutte le metodologie sviluppate sono verificate da simulazioni numeriche e test di laboratorio per entrambi i problemi. Difetti puntuali, di diversa gravità, sono stati indagati per la struttura a trave. Per il pannello quattro tipi diversi di cambiamento strutturale sono stati studiati, rimuovendo alcune viti per separare i rinforzi dal pannello, oppure rimuovendo un intero rinforzo. Le FRF dei pannelli in ogni stato sono stati registrati da due sistemi di misura (LMS Hammer Test e PSV-Laser Scanner Vibrometro) e per ogni scenario le relative forme modali sono state estratte da un algoritmo Polymax di LMS Software Testlab. Per simulare i disturbi sperimentati durante le prove sperimentali, i dati di misura generati da MSC/Nastran sono stati inquinati con un livello di 3% di rumore casuale. Questo studio mostra la potenzialità dei metodi proposti per un rilevamento semplice e rapido, ed una localizzazione precisa ed affidabile, del danno nelle strutture. Gli esempi riportati mostrano anche che alcuni metodi proposti (per esempio metodi basati sulla PCA e sulla trasmissibilità) sono altamente efficaci nel rilevare e localizzare i danni, permettendo il loro monitoraggio anche con dati inquinati da rumore di misura.
Vibration-based damage identification techniques
REZVANI, KAMAL
Abstract
Higher operational loads, greater complexity of design and longer lifetime periods imposed in civil, mechanical and aerospace structures, make it increasingly important to monitor the health of these structures. A wide variety of highly effective non-destructive methods, such as acoustic or ultrasonic methods, magnet field methods, penetrating liquids, eddy-current methods or thermal field methods, and so on, are currently available for the detection of defects. Unfortunately, they are all localized techniques, implying long and expensive inspection time; often, structural components are not inspected just because of their inaccessibility and damage can propagate to critical levels between the inspection intervals. The drawbacks of current inspection techniques have led engineers to investigate new methods for continuous monitoring and global condition assessment of structures. That is the case for methods based on vibration responses that allow one to obtain meaningful time and/or frequency domain data and calculate changes in the structural and modal properties, such as resonance frequencies, modal damping and mode shapes, FRF based methods, and use them with the objective of developing reliable techniques to detect, locate and quantify damage. The fundamental idea of the vibration methods for damage detection is that modal parameters (natural frequencies, mode shapes, and modal damping) are functions of the physical properties of the structure (mass, damping, and stiffness). Therefore, changes in the physical properties will cause detectable changes in the modal properties. So, this process involves the observation of a structure/system over time using periodical measurements. In other words, most vibration-based damage detection methods can be considered as some form of the pattern recognition problem as they look for the discrimination between two or more signal categories, e.g. before and after a structure is damaged or differences in the damage levels or locations. The work here presented addresses the subjects of damage detection, localization and quantification in structures. The reported examples show the implementation and comparison of a number of various damage detection and localization methods based on vibration responses changes. The objective of such a study is to ascertain the possibility of using various damage detection and localization methods with and without the need for modal identification. In recent years, the authors have developed some simplemethods and tools based upon the use of mode shapes and frequency response functions (FRFs), which seem promising and have given good results in some practical applications. The advantage of using such methods, which is based on mode shapes, is that only measured mode shapes are required in damage identification, without having to know the complete stiffness and mass matrices of the structure. To use of directly measured FRFs data, which provide an abundance of information, is further beneficial as the execution of experimental modal analysis is not needed, thus greatly reducing human induced errors. Here, all methods are first tested on data of a simple steel beam structure to assess their feasibility and performance. Then all proposed methods are applied to a more complicated structure, a typical aircraft stiffened panel, for extension and validation purposes. All developed methodologies are verified by numerical simulations and laboratory testing for both structure. As defects, notch type damage of different severity are investigated for the beam structure. For the panel structure four different types of structural change are studied. i.e. remove some screw for disconnecting the stiffener from the panel and create a saw cut along the all the depth of the hight of second stiffener. FRFs of panel in each state are recorded by two measurement systems (LMS Hammer Testing and PSV-Laser Scanner Vibrometer) and for each scenarios the related mode shapes are extracted by PolyMAX algorithm of LMS Testlab software. In addition, numerical simulated examples, after correlation with the real experimental one, of both mentioned structures will be used, as well as an experimental test cases, where damage is inflicted in a free-free condition. To simulate noise disturbances experienced during the experimental testing, for the numerical simulations of example beam, measurement data generated by MSC/Nastran are polluted with 3% of random noise level. The study shows the potential of the proposed methods for simple and rapid detection, accurate and reliable localization and monitoring of damage in structures. The reported examples also show that some proposed methods (i.e. PCA based, Transmissibility based) are highly capable for damage detection and localization and structural health monitoring even in the noise polluted data.File | Dimensione | Formato | |
---|---|---|---|
2015_03_PhD_Rezvani.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: PhD thesis Kamal Rezvani Areospace
Dimensione
90.75 MB
Formato
Adobe PDF
|
90.75 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/107294