Chloride penetration into silica fume concrete after 10 years of exposure in Aursundet Bridge

Aursundet Bridge which is a four-span and 486 m long cantilever bridge with a post-tensioned box girder, is located on the west coast of Norway. When the bridge was constructed in 1993-1995, special efforts with both the structural design and the selection of concrete quality were made in order to obtain a good durability. Thus, for the most exposed parts of the bridge, a type of concrete with 12.5% silica fume by weight of cement was selected. An ordinary portland cement and a silica fume with more than 90% SiO₂ were used, and the concrete mixture had a water/binder ratio of 0.36. In order to find out how such a high-volume silica fume concrete would be able to withstand chloride penetration from the local marine environment, detailed investigations of the bridge after 3 and 10 years of exposure were carried out. As part of the investigations, a number of concrete samples both for determination of chloride penetration and material characterization were removed. From the bridge columns, the samples for control of chloride penetration were partly taken from the splash zone at a level of 3.6 m above the mean water level and partly within the tidal zone having an extent of +/-1.43 m. The material characterization included the testing of degree of capillary suction, suction porosity and air porosity. These tests were carried out on "undisturbed" concrete samples taken from the bridge column by sawing 40-50 mm deep grooves about 20 mm apart on the concrete surface. Thus, upon arrival in the laboratory, it was possible to determine the moisture content in the undisturbed samples. After 10 years, accelerated chloride diffusivities were also determined by use of an accelerated migration test method. In addition, a control of compressive strength was carried out. Both after 3 and 10 years of exposure, a significantly different chloride penetration between the windward and the leeward side of the bridge columns had taken place, and this is a pattern of chloride penetration which is typically observed on concrete bridges along the Norwegian coastline. For the most protected parts of concrete coastal bridges, experience shows that the salt accumulates, while for the most exposed parts, the intermittent rain is successively washing off the salt from the concrete surface. For the high silica fume containing concrete, a very high resistance against chloride penetration was observed. Thus, after 10 years of exposure in the splash zone and the tidal zone, apparent diffusion coefficients of 5.3 and 3.3 × 10^-13 m²/s were observed, respectively, and based on the difference in chloride penetration after 3 and 10 years, a typical coefficient of 0.26 for the time dependence of the chloride diffusivity was obtained. The material characterization also confirmed the high quality of the bridge concrete, with suction and air porosities typically varying from 11.8 to 15% and from 1.6 to 3.2%, respectively. The observed values for accelerated chloride diffusivity and electrical resistivity of 5.0 × 10-¹² m ²/s and 234 ohm·m, respectively, also reflect the high resistance of the concrete against chloride penetration. By visual observations, the high silica fume containing concrete did not show any cracking of the concrete surface, but the test results reflect a high scatter and variability of the concrete quality, which is often observed for the testing of in situ concrete quality of existing concrete structures. The concrete showed a high moisture content with values typically varying from 86.2 to 88.7%, but a value of 75% was also observed. The combination of a concrete with a high resistance against chloride penetration and a nominal concrete cover of 100 mm had given fhe concrete bridge a high safety against steel corrosion. Thus, on the basis of a probability-based durability analysis, a probability for steel corrosion of approximately 8% after a period of 100 years was obtained.