Experimental and prediction model of axial compressive responses of corroded RC cylinders strengthened with CFRP

Currently, the capacity of corroded RC cylinders has been proven to significantly improve after being strengthened with carbon fiber reinforced polymer (CFRP). However, existing experimental data remains insufficient, resulting in few models available for predicting the post-strengthening capacity of corroded cylinder. The current understanding of the mechanism by which corrosion affects the mechanical properties of CFRP strengthened cylinders is also not yet clear. Against this background, this study conducted monotonic axial compression tests on CFRP-strengthened corroded RC cylinders. The experimental results indicate that while the CFRP layer predominantly controls the specimen's load-bearing capacity, the reduction in CFRP rupture strain due to increased corrosion rates should not be overlooked. When the corrosion rate reaches 35 %, the experimentally measured average rupture strain was only 65 % of the theoretical value, marking a 30 % decrease compared to non-corroded specimens. Subsequently, a predictive model for peak stress and strain of CFRPconfined corroded RC cylinders was proposed, which considers the confinement mechanisms of stirrups and CFRP as well as the degradation of CFRP rupture strain. Leveraging the existing test data, a dataset comprising 80 CFRP-strengthened corroded RC cylinder specimens was established to calibrate the key parameters in the proposed prediction model. Comparative analysis with existing models indicates that the peak stress and strain prediction model proposed in this study demonstrated the highest forecasting accuracy (the average errors for peak stress and strain were 14 % and 32.6 %, respectively), while existing other predictive models based on limited datasets suffer from a lack of universality. The experimental data and research conclusions of this article can provide reference for engineering practice and the establishment of models in subsequent research.