A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Updating a finite element (FE) model of a structure using test data is a necessary step for its experimental damage detection. Modeling structures with L-shaped beams and bolted joint connections for damage detection is a challenging task because the theoretical natural frequencies and mode shapes need to match with their test data for both the undamaged and damaged structures. The bending and torsional vibrations of a structure with L-beams can also be coupled in most of their modes. The updated FE models need to capture this behavior within an acceptable model size. Three physics-based parameterization methods are developed in this work: 1) a simple stiffness and mass updating method using a solid element model; 2) an inertia and geometric offset updating method for L-beams using a combined beam and shell element model, and a stiffness or geometric parameter updating model for bolted joints using a solid element model; and 3) a geometric parameter updating method for L-beams using a combined solid and shell element model, and a stiffness or geometric parameter updating model for bolted joints using a solid element model. The first method can model a L-shaped beam accurately and has only one parameter to update in the procedure of modeling a bolted joint connection, although the degrees of freedom are large. The second method has the advantage of reducing dramatically the model size without losing the model accuracy. The number of the updated parameters is increased for the L-beam and a more detailed model for bolted joints is employed. The third method contains far fewer degrees of freedom than the first method and the number of the updated parameters for the L-beam is less than that in the second method. More importantly, the updated parameters for the L-shaped beam have more physical meaning. For each method the errors between the measured and calculated natural frequencies of the structures are within 2% for at least the first 11 modes considered, and the corresponding measured and calculated mode shapes also match, with the modal assurance criterion numbers over 94%.
Updating a finite element (FE) model of a structure using test data is a necessary step for its experimental damage detection. Modeling structures with L-shaped beams and bolted joint connections for damage detection is a challenging task because the theoretical natural frequencies and mode shapes need to match with their test data for both the undamaged and damaged structures. The bending and torsional vibrations of a structure with L-beams can also be coupled in most of their modes. The updated FE models need to capture this behavior within an acceptable model size. Three physics-based parameterization methods are developed in this work: 1) a simple stiffness and mass updating method using a solid element model; 2) an inertia and geometric offset updating method for L-beams using a combined beam and shell element model, and a stiffness or geometric parameter updating model for bolted joints using a solid element model; and 3) a geometric parameter updating method for L-beams using a combined solid and shell element model, and a stiffness or geometric parameter updating model for bolted joints using a solid element model. The first method can model a L-shaped beam accurately and has only one parameter to update in the procedure of modeling a bolted joint connection, although the degrees of freedom are large. The second method has the advantage of reducing dramatically the model size without losing the model accuracy. The number of the updated parameters is increased for the L-beam and a more detailed model for bolted joints is employed. The third method contains far fewer degrees of freedom than the first method and the number of the updated parameters for the L-beam is less than that in the second method. More importantly, the updated parameters for the L-shaped beam have more physical meaning. For each method the errors between the measured and calculated natural frequencies of the structures are within 2% for at least the first 11 modes considered, and the corresponding measured and calculated mode shapes also match, with the modal assurance criterion numbers over 94%.
Finite element modeling of structures with L-shaped beams and bolted joint connections for vibration analysis and damage detection
Finite-Elemente-Modellierung von Strukturen mit L-Trägern und Bolzenverbindungen für die Schwingungsanalyse und Schadenerkennung
2007
25 Seiten, 19 Bilder, 19 Tabellen, 23 Quellen
Conference paper
English
Finite Element Modeling for Static and Dynamic Analysis of Structures with Bolted Joint
| British Library Online Contents | 2006
Finite element analysis of bolted connections for PFRP composites
| British Library Online Contents | 1996