Shear behaviour of CFRP-prestressed, steel-reinforced concrete girders
Author: Cyrill Habegger
Language: English
Abstract
Many reinforced concrete structures in the Swiss railway system are reaching their end of service life and therefore, they have to be replaced by new structures. To reduce future maintenance and replacement efforts, a system solution for short-span railway bridges with a focus on maximum durability is currently under development.
To achieve this goal, prestressed Carbon Fibre Reinforced Polymer (CFRP) reinforcement is used in combination with stainless steel and Ultra High Performance Concrete (UHPC). The use of prestressed CFRP bars as reinforcement is a new approach and presents a scientific knowledge gap. To improve understanding of the shear behaviour of the longitudinal girders in the prototype bridge, experimental testing and further analysis are carried out.
The specimen includes 12 prestressed CFRP bars and 6 stainless steel bars as longitudinal reinforcement. The stirrups are spaced at a distance of 150mm. The girder consists of a T-shaped cross-section with a total height of 550mm and a flange width of 500 mm. Fibre Optic Sensor (FOS) and Digital Image Correlation (DIC) measurement systems are used to investigate the behaviour. The strain distribution followed the expected trends.
The lower the reinforcement bars in the cross-section, the higher the measured strains in the longitudinal reinforcement. The stirrups were activated at a specific load when a crack crossed them. The experiments showed for both tests at first a bending failure initiated by concrete crushing in the flange as a result of high compressive stresses from bending.
The second test was loaded to ultimate failure and resulted in a shear failure in the web. The strains from the stirrup data indicated that they did not fully yield. Therefore, the most likely failure is due to local concrete crushing in the web. The shear force transfer across a crack showed that the contribution by the stirrups is only about 70% and 55%, respectively. The contribution of aggregate interlock and dowel action is indicated as minor.
Therefore, another contributor is transferring load across the cracks. The strut-and-tie model and the cross-sectional analysis showed that the resistance capacity is underestimated by these approaches. Additionally, the strain-based cross-section calculations aligned better for smaller loads.
The combination of the prestressed CFRP bars with stainless steel and UHPC showed good performance. The CFRP bars were fully utilised but they did not fail during the test. The interpretation is limited by the unknown DIC accuracy. Due to the missing statistical parameters, the interpretations were mainly qualitative. In addition, some FOS strains showed unreliable results at high strains.
This thesis contributes to a better understanding of structures with CFRP reinforcement under shear load. These insights are helpful for the further research on the short-span railway bridge.
