Behaviour of bond between CFRP sheet and steel under impact tensile loading

The bond between steel and CFRP composites is a critical issue for CFRP strengthening and/or repairing technique. The strengthened and/or repaired structures are likely to be subjected to dynamic loadings in practice. However, the bond between CFRP composites and steel structures is not well understood, particularly under dynamic tensile loadings. This thesis reports an investigation on the influence of impact loading on the bond between CFRP sheet and steel. A series of impact tensile tests were carried out to investigate the loading speed effect on the tensile mechanical properties of composites of CFRP sheet and two adhesives, Araldite 420 and MBrace saturant, and steel plates. The loading speed effect on the shear strength of Araldite 420 adhesive was also studied. In addition, empirical equations were proposed, which can determine the tensile properties of CFRP sheet, Araldite 420 and MBrace saturant adhesives as a function of strain rate within the range considered. The experimental results indicated that the loading speed affects the tensile properties and failure modes of the CFRP sheet and both adhesives. The proposed empirical equations provide best fit to the experimental data. The yield and ultimate stresses of steel plate are less influenced by loading speed compared to the shear strength of Araldite 420, which markedly increases by approximately 3 times the quasi-static shear strength. A further two sets of experiments were conducted on CFRP/steel double-strap joints with 1 and 3 CFRP layers at quasi-static loading and dynamic loading speeds of 3.35 m/s, 4.43 m/s and 5 m/s to study the loading speed effect on the bond between CFRP sheet and steel plate. The influence of the adhesive on the dynamic behaviour of the adhesively-bonded joints was also shown. One set of samples was prepared using MBrace saturant, whereas the other used Araldite 420. Samples of both sets were of various bond lengths ranging from 10 to 100 mm. The parameters studied were the bond strength, effective bond length, failure mode and strain distribution along the bond length. The experimental results showed that the bond strength of both joints with 1 and 3 CFRP layers prepared utilising both adhesives is influenced by loading speed, although the general trends differ. Increasing the number of CFRP layers insignificantly influences the quasi-static bond strength of both Araldite 420 and MBrace saturant adhesives. Conversely, increasing the number of CFRP layers markedly affects the dynamic bond strength. Loading speed considerably affects the failure patterns of both joints prepared using both adhesives, particularly when increasing the loading speed from quasi-static to 3.35 m/s, due to the shear strength enhancement of the adhesives. The effective bond length for joints prepared using both adhesives is slightly influenced by the loading speed and their general trends at dynamic loadings differ. Identical strain distribution profiles with only slight differences in the measured strain values are found for joints with 1 CFRP layer for both adhesives, while a clear difference is observed in the strain distributions of joints with 3 CFRP layers for both adhesives. An additional set of experiments was conducted to examine the effect of direct impact tensile loading on the bond strength of the adhesion between CFRP sheet and steel using pull-off tests. Special CFRP/steel samples, prepared with 1 and 3 CFRP layers using both MBrace saturant and Araldite 420 adhesives, were tested at quasi-static and dynamic loading speeds of 3.35, 4.43 and 5 m/s. It was found the bond strength of adhesion for both joints with 1 and 3 CFRP layers and for both adhesives increases overall by about 100% at the dynamic loading speeds considered. The values for samples with 3 CFRP layers are lower than those from samples with 1 CFRP layer for both adhesives. Considerable changes in the dynamic failure patterns compared to the quasi-static mode are found for joints with 1 CFRP layer and for both adhesives, whereas the failure mode for samples with 3 CFRP layers using both adhesives remain CFRP delamination through the CFRP layers. The dynamic behaviour of the bond between CFRP sheet and steel plate under dynamic loading was further studied using the finite element method. The individual components of the double-strap joints, the CFRP sheet, adhesive and steel plate, were first numerically analysed under quasi-static and dynamic loading speeds and validated against the experimental results. Both types of CFRP/double-strap joints with 1 and 3 CFRP layers using Araldite 420 were then numerically analysed at the same loading speeds as those used in the experimental program. In the numerical validation of the double-strap joints, finite element predictions of the ultimate failure load, effective bond length, failure patterns and strain distributions along the bond length were compared with the experimental results. The numerically-predicted and experimentally-obtained ultimate failure loads, effective bond lengths, failure patterns and strain distributions along the bond length of both joints correlated reasonably well. Finally, some recommendations for the application of CFRP sheets to strengthen and/or rehabilitate steel structures are provided, and possible future studies of the dynamic behaviour of CFRP-composite-reinforced metal structures are identified.