This paper presents an experimental evaluation of the fire resistance of glued-in rod timber joints using epoxy resin, with and without modification. A heat-resistant modified resin was designed by adding inorganic additives into the epoxy resin, aiming to improve the heat resistance. Joints that were made using the modified epoxy resin at room temperature showed a bearing capacity comparable to those with commercial epoxy resin. Twenty-one joint specimens with the modified epoxy resin and six with a commercial epoxy resin were tested in a fire furnace to evaluate the fire resistance. The main failure mode was the pull-out of the rod, which is typical in fire tests of this type of joints. As to the effects of the test parameters, this study considered the effects of adhesive types, sectional sizes, stress levels, and fireproof coatings. The test results showed that the fire resistance period of a joint can be evidently improved by modifying the resin and using the fireproof coating, as the improvements reached 73% and 35%, respectively, compared with the joint specimens with commercial epoxy resin. It was also found that, for all specimens, the fire resistance period decreased with an increase in the stress level and increased with an increase in the sectional sizes.
This paper presents the results of long-term experiments performed on three timber-concrete composite (TCC) beams. An innovative fabricated steel plate connection system, which consists of screws and steel plates embedded in concrete slabs, was adopted in the TCC beam specimens. The adopted shear connection can provide dry-type connection for TCC beams. Steel plates were embedded in concrete slabs while the concrete slab was constructed in factories. The timber beam and concrete slab can be assembled together using screws at the construction site. In this experimental programme, the beam specimens were subjected to constant loading for 613 days in indoor uncontrolled environments. The influence of long-term loading levels and the number of shear connections on the long-term performance of TCC beams was investigated and discussed. The mid-span deflection, timber strain, and interface relative slip at the positions of both connections and beam-ends were recorded throughout the long-term tests. It was found the long-term deflection of the TCC beam increased by approximately 60% while the long-term loads were doubled. Under the influence of the variable temperature and humidity, the TCC specimens with 8 shear connections showed slighter fluctuations compared with the TCC beam with 6 shear connections. In the 613-day observation period, the maximum deflection increment recorded was 6.56 mm for the specimen with eight shear connections and 20% loading level. A rheological model consisting of two Kelvin bodies was employed to fit the curves of creep coefficients. The final deflections predicted of all specimens at the end of 50-year service life were 2.1~2.7 times the initial deflections caused by the applied loads. All beam specimens showed relative small increments in mid-span deflection, strain and relative slip over time without any degradations, demonstrating the excellent long-term performance of TCC beams using the innovative steel plate connection system, which is also easily fabricated.