Real-Time Structural Health Monitoring of Vibrating Systems

Description

Targeted at researchers and practitioners in the field of science and engineering, the book provides an introduction to real time structural health monitoring. Most work to date is based on algorithms that require windowing of the accumulated data, this work presents a coherent transition from the traditional batch mode practice to a recently developed array of recursive approaches. The book mainly focuses on the theoretical development and engineering applications of algorithms that are based on first order perturbation (FOP) techniques. The development of real time algorithms aimed at identifying the structural systems and the inflicted damage, online, through theoretical approaches paves the way for an in-depth understanding of the discussed topics. It then continues to demonstrate the solution to a class of inverse dynamic problems through numerically simulated systems. Extensive theoretical derivations supported by mathematical formulations, pivoted around the simple concepts of eigenspace updates, forms the key cornerstone of the book. The output response streaming in real time from multi degree of freedom systems provide key information about the system’s health that is subsequently utilized to identify the modal parameters and the damage, in real time. Damage indicators connotative of the nature, instant and location of damage, identified in a single framework are developed in the light of real time damage case studies. Backed by a comprehensive assortment of experimental test-beds, this book includes demonstrations to emulate real life damage scenarios under controlled laboratory conditions. Applicability of the proposed recursive methods towards practical problems demonstrate their robustness as viable candidates for real time structural health monitoring. Cover Title Page Copyright Page Preface Acknowledgements Table of Contents Symbols Part: I 1. Introduction 1.1 Need for SHM development 1.2 Definitions 1.3 Components of SHM 1.4 Need for real-time SHM 1.5 Motivation for this book 1.6 Organization of the book References 2. Mathematical Preliminaries 2.1 Deterministic and random signals 2.2 Fourier transform 2.3 Frequency response function 2.4 Impulse response function 2.5 Power spectral density 2.6 Eigenperturbation–a bird’s eye view 2.7 Recursive covariance estimation in a higher order perturbation framework 2.8 A note on eigenvalue sensitivity 2.9 Gershgorin’s theorem 2.10 Time series analysis 2.11 Summary References 3. Single-sensor Real Time Damage Detection Techniques: RSSA and its Variants 3.1 Introduction 3.2 Background 3.2.1 Singular Spectrum Analysis (SSA) 3.2.2 Multichannel Singular Spectrum Analysis (MSSA) 3.2.3 RPCA in structural dynamics 3.3 RMSSA-RPCA adaptive algorithm 3.4 Framework of the RMSSA-RPCA algorithm 3.5 Numerical models 3.5.1 5-DOF linear structural system 3.5.2 5-DOF Base-isolated structural system 3.6 Damage identification results 3.6.1 RSSA algorithm results: 5-DOF BW-system 3.6.2 5-DOF linear system for RMSSA-RPCA algorithm 3.6.3 5-DOF Base-isolated system 3.6.4 Threshold for damage detectability of RMSSA-RPCA algorithm 3.7 Robustness of the RMSSA-RPCA algorithm 3.7.1 Damage detection in the presence of low frequency-operational noise 3.7.2 Case study for the 5-DOF base isolated system using non-stationary excitation 3.7.3 Comparison with RPCA algorithm 3.8 Experimental verification 3.8.1 Description of the setup and measurement program 3.8.2 Results 3.9 Conclusions References 4. Multi-sensor Real Time Damage Detection Techniques (A): RPCA 4.1 Motivation 4.2 Problem formulation 4.3 Recursive covariance estimation and FOEP 4.3.1 RPCA: Theoretical development using POMs 4.4 Real time condition indicators 4.4.1 Recursive residual error (RRE) 4.4.1.1 Temporal RRE 4.4.1.2 Local RRE 4.4.2 Recursive eigen vector change 4.4.3 Outlier detection using correlation coefficient (ρ) 4.5 Proposed algorithm 4.6 Numerical example 4.6.1 Structural model and simulation parameters 4.6.2 Results for white noise input excitation 4.6.2.1 Temporal damage detection results 4.6.2.2 Spatial damage detection results 4.6.3 Results for underdetermined case—white noise excitation 4.6.4 Comparative study with batch PCA 4.6.5 Results for El Centro ground excitation References 5. Multi-sensor Real Time Damage Detection Techniques (B): RPCA-TVAR 5.1 Motivation 5.2 Background 5.3 RPCA and structural dynamics: A POC based formulation 5.4 TVAR modeling 5.5 Damage sensitive features 5.5.1 Time varying auto-regressive coefficients 5.5.2 Recursive statistics on TVAR coefficients 5.6 Proposed algorithm 5.7 Numerical example 5.8 Temporal damage detection results 5.8.1 Spatial damage detection results 5.8.2 Results for El Centro ground excitation 5.8.3 Results for time-diluted damage 5.8.4 Results for underdetermined case-white noise excitation 5.9 Experimental study References 6. Multi-sensor Real Time Damage Detection Techniques (B): Real Time Structural Damage Detection Using Recursive Canonical Correlation Analysis 6.1 Motivation 6.2 Background 6.3 RCCA: Detailed derivation 6.4 Recursive damage sensitive features 6.5 Proposed algorithm 6.6 Detection results using proposed algorithm 6.6.1 Temporal damage detection studies for the B-W systems excited using white noise 6.6.1.1 Temporal damage detection results for the 5 DOF B-W system 6.6.1.2 Spatial damage detection results for the Bouc-Wen 5 dof model 6.6.2 Detection results for the 5 DOF B-W system excited by El Centro ground motion 6.6.3 Case study for an under determined system 6.6.4 Case study of a 7 DOF B-W system excited using white noise 6.6.5 A note on the effect of the increased DOF on the detection results using the proposed algorithm 6.6.6 Case study for a strongly nonlinear system: A 5 DOF structure modeled with Duffing oscillator on its 3rd floor 6.6.7 A case study with negligible non-linearity: Spatial damage detection 6.7 Performance check of the proposed method against RPCA: A comparative study 6.7.1 Case studies for the 5 DOF B-W system using white noise 6.7.1.1 Case study for 15% global damage detection using 5 DOF B-W system 6.7.1.2 Case study for 25% local damage detection using 5 DOF B-W system 6.7.2 Case study for the 5 DOF B-W system excited using El Centro ground motion 6.8 Experimental verifications 6.8.1 Detection results for the experimental trials: A 164comparative study with RPCA References 7. Real-time Modal Identification— RCCA Based Approach 7.1 Motivation 7.2 Background 7.2.1 BSS based formulation of CCA 7.2.1.1 PCA and structural dynamics 7.2.2 Recursive covariance estimation and FOEP 7.3 RCCA: Detailed derivation 7.4 Proposed algorithm 7.5 Model description 7.5.1 5 DOF mass, spring and dashpot system: Model description 7.5.2 5 DOF B-W model description 7.6 Modal identification results 7.6.1 Modal identification results using batch CCA 7.6.2 Modal identification results using RCCA 7.6.2.1 Recursive modal identification results for 5–DOF linear system 7.7 Real time damage detection: An application of RCCA 7.7.1 Recursive damage sensitive features 7.7.2 Damage detection results for linear 5–DOF system 7.7.3 Damage detection results for 5 DOF B-W system using RCCA 7.7.3.1 Global damage detection results using the 5 DOF B-W system 7.7.3.2 Local damage detection results using the 5 DOF B-W system 7.8 Experimental verifications 7.8.1 2 DOF shear building experimental setup References Part: II 8. Recursive SHM Practical Applications 8.1 Introduction 8.2 UCLA factor building—A general overview 8.3 UCLAFB case study—RPCA-RRE 8.4 UCLAFB case study—RPCA-TVAR 8.5 UCLAFB case study—RSSA 8.6 Case study for the IASC–ASCE benchmark structure 8.7 Phase-I ASCE-SHM benchmark—A general overview References 9. Conclusions and Future Prospects 9.1 Summary 9.2 Overview of Results and Contributions 9.3 Future Prospects of Real-Time Monitoring 9.4 Conclusions References Index