Study On Damage Detection And Model Updating For Offshore Platform Structures

Offshore structures continually accumulate damages as a result of the action ofvarious environmental forces during their service life. Clearly, the development ofrobust techniques for damage detection is crucial to avoid the possible occurrence ofa catastrophic structural failure. And model updating is of great importance since itprovides accurate Finite Element Model (FEM) for dynamic analysis, modal testingand damage detection, etc.Damages in offshore structures would induce the changing of modal parameters;and damage detection algorithms based on modal parameters alteration are commonlyrecognized to be one of most potential technologies. Damage detection and modelupdating require measured modal data from modal testing. In the modal testing ofoffshore structures, due to limited sensors, difficulty and inaccuracy in measuring ro-tational Degree-of-Freedoms (DOFs) and underwater DOFs, the number of measuredcoordinates are far less than the DOFs of the FEM, namely Spatially Incompleteness(SI) of the measured modal shapes. The SI issue proposes a more severe challengeto damage detection and model updating algorithms. And this thesis performed threeinnovative studies as followed.In dealing with Spatial Incompleteness (SI), CMCM in conjunction with Guyanscheme based transformation matrix iterative updating method is proposed. Specif-ically, either Guyan (static condensation) or SEREP (System Equivalent ReductionExpansion Process) transformation matrix, between the master and slave DOFs, isemployed in the model reduction or modal expansion process. One theoretical devel-opment is an iterative procedure to compute the transformation matrix associated withthe (unknown) damaged structure.Different from the concept of model reduction or modal expansion, an new it-erative mathematical approach, named as Measured Model Salve DOFs Updating(MMSDU) method, is proposed in dealing with the SI issue. A 5-DOFs mass-spring system is studied for verifying the accuracy and effectiveness of the proposedMMSDU method, also for demonstrating the detailed procedure; and a 3-D 72-DOFsoffshore jacket platform structural model to demonstrate the feasibility and capabil-ity of the MMSDU method for complex structures. One important improvement isthat the MMSDU method can update models and detect damages by employing onlyfrequency data.A new sensitivity analysis index for individual member of complex structures is proposed. And an exsiting offshore platform is taken as the example to carry out thesensitivity analysis of individual members, and Singular Value Decomposition (SVD)of the CMCM core matrix. The obtained results optimized the choice of the modelupdating parameters.The CMCM method was applied to the model updating and damage detectionin offshore structures, the proposed CMCM with Guyan reduction iterative updatingmethod and MMSDU method resolve the SI issue at different levels creatively. Mainconclusions are drawn: (1) the CMCM method can accurately update mass and stiff-ness of jacket structures, by employing a few lower-order spatial complete modes, noneed to pair or scale, and the updated model keeps its physical connection; (2) theCMCM method can diagnose the existence of damages, localize damages and esti-mate damage severities simultaneously by employing a few lower-order modes; (3)the proposed CMCM with Guyan reduction transformation matrix updating methodresolves the SI issue at a certain level; (4) the proposed MMSDU method resolvesthe SI issue further, and the MMSDU method can update models and detect damagesby using only frequency data; (5) Sensitivity analysis of individual member guidesthe choice of model updating parameters, and reasonable updating parameters choiceimproves the updating results.