In design of bridges, restraint effects such as thermal actions and shrinkage must be included. One of the thermal load cases given in Eurocode describes differences in temperature between structural parts. When the load case is applied to portal frame bridges, large stresses can appear in the transversal direction of the structure, along the rigid connections between the structural parts. The load description is however very simplified, which may lead to an overestimation of the resulting stresses in the structure. Furthermore, the restraint effects are overestimated even further if the reduction of restraint when cracking occurs is not included in design. However, there is no commonly accepted methodology of doing this in design of the structural type today.
The aim of the project is the develop a more realistic thermal load case for design of portal frame bridges in Sweden. A model for thermal simulation of bridges by use of weather data was validated using measurements of temperature in a portal frame bridge during one year, and was thereafter used to develop a new load case. Also, non-linear finite element analyses were made to investigate the effect of cracking on the crack widths. The finite element model was validated by comparing analysis results with test results of cracking in base restrained walls performed in previous research. The analyses of cracking in portal frame bridges showed that the crack widths become smaller than what is estimated if the effect of stiffness reduction due to cracking is disregarded, and that the reinforcement amount has a smaller influence on the crack widths. More research is however needed in order to suggest a method for how to include the effect of cracking on the restraint stresses and resulting crack widths in design situations.