| dc.description.abstract | The aim of this thesis is to gain a broader understanding of bridge modeling and nonlinear analyses of structures. Røydalen bridge has been chosen as a case study and the modeling and analyses of the bridge has been performed in Sofistik. Material properties, cross sections and loads will be defined in the model. The model will then be used to perform linear and nonlinear analyses with design limit state load combinations using finite element method. For the linear analyses the loads will be combined using the principle of superposition. As this is not applicable for the nonlinear analyses, separate analyses for each load combination will be executed.
Structures are usually designed based on linear analyses, but since linear analyses are nonlinear approximations they might be designed conservatively. Difference in forces for linear and nonlinear analyses will be investigated and compared to establish if the linear analyses provide too rigid models, hence conservative results.
Nonlinear analyses are rarely used as they are time consuming and the required expertise from the analyst is quite high. It will be investigated if a shorter method can be used and whether this method can be an addition to the linear analyses. The method involves using only the most unfavorable load combinations from the linear analyses in the nonlinear analyses.
Six full nonlinear and two limited nonlinear analyses was performed with load combinations in ultimate limit state (ULS) and linear analyses for both ULS and serviceability limit state. The results from the full nonlinear analyses supports the assumption that linear analyses are to some extent conservative. The results from the limited nonlinear analyses are promising but the analyses method requires further investigation. | |
