【博士論文】学術データベース

博士論文 / AN ENERGY-BASED DAMPING EVALUATION USING BAYESIAN MODEL UPDATING FOR VIBRATION-BASED STRUCTURAL HEALTH MONITORING OF STEEL TRUSS BRIDGES 鋼トラス橋ヘルスモニタリングのためのベイズ推定によるモデルアップデートを利用したエネルギー的振動減衰評価法

著者

書誌事項

タイトル

AN ENERGY-BASED DAMPING EVALUATION USING BAYESIAN MODEL UPDATING FOR VIBRATION-BASED STRUCTURAL HEALTH MONITORING OF STEEL TRUSS BRIDGES

タイトル別名

鋼トラス橋ヘルスモニタリングのためのベイズ推定によるモデルアップデートを利用したエネルギー的振動減衰評価法

著者名

SAMIM MUSTAFA

学位授与大学

埼玉大学 (大学ID:0019) (CAT機関ID:KI00020X)

取得学位

博士(学術)

学位授与番号

乙第241号

学位授与年月日

2017-03-22

注記・抄録

The existing infrastructure such as bridges which are the valuable national assets for transportation and economy are required to be maintained properly to ensure the performance and condition for their continuous operation. Difficulties in practical application of vibration-based structural health monitoring (SHM) of structures include considerable amount of uncertainties in structural modeling and vibration measurement and sensitivity issues of modal parameters due to local damage in case of large structure. This dissertation proposed an analytical framework for SHM addressing the aforementioned difficulties by combining two techniques: A Bayesian based probabilistic approach for finite element model (FE-model) updating that accounts for the underlying uncertainties and an energy-based damping model for detecting damage at local level using a small number of sensors.An efficient and robust Bayesian model updating was presented in this dissertation by introducing a new objective function and a realistic parameterization of mass and stiffness matrices. In this framework, the likelihood function for mode shapes was formulated based on the cosine of the angle between the analytical and measured mode shapes which does not require any scaling or normalization as compared to conventional Bayesian methods. Four stiffness parameters were introduced for each element considering both sectional and material properties to take into account variation in each element due to local damage. The proposed updating method was validated experimentally by updating a FE-model of existing steel truss bridge utilizing the vibration data obtained from limited number of sensors by a car running test.It has been recognized that the damping is more sensitive to local damage and the advantage of using damping is that the damping change in global modes affected by local damage can be identified with a small number of sensors. In this dissertation, an energy-based damping model was introduced for practical and effective SHM by estimating the contribution of modal damping ratios from different structural elements utilizing the data from updated FE-model and the identification results of damping from a small number of sensors. A previous study reported that the studied bridge with damage at local diagonal member showed a significant increase in the damping of global vibration mode of the structure. The present study utilized the energy-based damping evaluation to identify possible cause of the modal damping increase by observing the change in the contribution from different structural elements on the modal damping ratios.

Title Page iAcknowledgement vAbstract viiTable of Contents ixList of Figures xiiiList of Tables xvNomenclature xviiAbbreviations xixCHAPTER 1 INTRODUCTION 11.1 Research Background and Motivation 11.2 Objectives of the Study 31.3 Outline of the Dissertation 4CHAPTER 2 LITERATURE REVIEW 72.1 An Overview of Vibration-based Structural Health Monitoring 72.2 Vibration-based Structural Health Monitoring Techniques 9 2.2.1 Non-model based Methods 9 2.2.2 Model-based Inverse Methods 102.3 Difficulties in Practical Application of Vibration-based SHM 12 2.3.1 Issues with Sensing and System Identification 13 2.3.2 Issues with Model-based SHM 14 2.3.3 Issues with Sensitivity of Modal Parameters to Local Damage 152.4 Proposed Framework to Overcome Aforementioned Difficulties 15CHAPTER 3 BAYESIAN PROBABILISTIC APPROACH FOR MODEL UPDATING USING LIMITED SENSOR DATA 193.1 Introduction 193.2 Application of Bayesian Probabilistic Approach to FE-model Updating 19 3.2.1 Bayesian Probabilistic Approach 19 3.2.2 Parameterization of Mass and Stiffness Matrices 20 3.2.3 Formulation of Likelihood Function 20 3.2.4 Formulation of Eigenvalue Equation Errors 22 3.2.5 Formulation of Prior PDF 22 3.2.6 Formulation of Posterior PDF 233.3 Optimal Parameter Vectors 23 3.3.1 Optimization for Auxiliary Variables and Lagrange Multipliers 24 3.3.2 Optimization for Mode Shape Vectors 24 3.3.3 Optimization for Frequencies 24 3.3.4 Optimization for Stiffness Parameters 24 3.3.5 Optimization for Mass Parameters 253.4 Studied Steel Truss Bridge 25 3.4.1 Test Structure Description 25 3.4.2 Finite Element Model of Test Structure 253.5 Experimental Validation of Proposed Updating Framework 27 3.5.1 Vibration Measurements and System Identification 27 3.5.2 Parameterization of Mass and Stiffness Matrices for the Truss Bridge 30 3.5.3 Model Updating Results and Discussion 313.6 Conclusions 34CHAPTER 4 ENERGY-BASED DAMPING EVALUATION FOR SHM 374.1 Introduction 374.2 Energy-based Damping Evaluation 37 4.2.1 Energy-based Damping Model for Test Bridge 38 4.2.2 Elemental Damping Evaluation 40 4.2.3 Evaluation of Modal Energies 404.3 Application to Test Bridge 41 4.3.1 Experimental Damping Identification 41 4.3.2 Identification of Loss Factors and Friction Coefficients 42 4.3.3 Evaluation of Analytical Modal Damping Ratios 44 4.3.4 Contribution of Modal Damping from Each Element 474.4 Conclusions 48CHAPTER 5 APPLICATION OF PROPOSED FRAMEWORK TO SHM WITH A SMALL NUMBER OF SENSORS 515.1 Introduction 515.2 Problem Description 515.3 Identification of Loss Factors for Damaged Span 545.4 Damage Detection Using Change in Analytical Modal Damping Ratios 55 5.4.1 Evaluation of Analytical Modal Ratios for Damaged Span 55 5.4.2 Re-analysis of Loss Factors for Damaged Span 57 5.4.3 Justification Against Change in Loss Factors and Discussion 585.5 Conclusions 62CHAPTER 6 SUMMARY AND FUTURE WORKS 636.1 Summary of the Contribution Made 636.2 Suggestions for Further Work 65REFERENCES 68APPENDIX A OPTIMAL SENSOR PLACEMENT FOR AN EXISTING STEEL TRUSS BRIDGE 73A.1 Introduction 73A.2 OSP Techniques 73 A.2.1 Effective Independence Method 74 A.2.2 Energy Matrix Rank Optimization 75 A.2.3 Modal Approach Using System Norms 76A.3 Results of OSP for Test Structure 77A.4 Conclusions 78References 79APPENDIX B DAMAGE DETECTION BY FE-MODEL UPDATING 81B.1 Application to SHM 81B.2 Damage Detection 82 B.2.1 Considering Simulated Damage 82 B.2.2 Considering Experimental Data from Damaged Span 85B.3 Conclusions 88APPENDIX C FORMULATION AND PARAMETERIZATION OF MASS AND STIFFNESS MATRICES 91C.1 Stiffness Formulation 91C.2 Mass Formulation 92C.3 Transformation of Coordinates 93C.4 Parameterization of Mass and Stiffness Matrices 94

主指導教員 : 松本泰尚

博士の専攻分野の名称 : 博士(学術)学位授与年月日 : 平成29年3月22日

各種コード

NII論文ID(NAID)

500001054963

NII著者ID(NRID)
  • 8000001173390
本文言語コード

eng

データ提供元

機関リポジトリ / NDLデジタルコレクション

外部リンク

博士論文 / 埼玉大学 / 学術

博士論文 / 埼玉大学

博士論文 / 学術

関連著者

博士論文 / 大学

博士論文 / 学位