Cyclic cumulative damage and precursor response characteristics of carbon-reinforced concrete lining in abandoned mine gas storage chambers

The surrounding rock-lining structure is highly susceptible to cyclic cumulative damage under injection-production cyclic stress. To address the insufficient resistance and damage perception of traditional lining materials under graded cyclic loading-unloading, a multi-carbon reinforced concrete (MCRC) was developed in this study. A rate-coupled graded cyclic loading-unloading test method was employed to investigate the cyclic damage evolution and precursor response characteristics of MCRC, integrating acoustic emission (AE), resistivity, and digital image correlation (DIC) techniques. The results demonstrate that incorporating carbon-based materials significantly enhances the mechanical performance and cyclic load resistance of concrete, with the compressive strength of the XH9 group increasing by 18.9%. The residual strain and elastic modulus exhibit a distinct three-stage evolution, and the carbon-based composites show superior crack inhibition and deformation compatibility compared with the control specimens. The carbon-based materials primarily refine the microstructure by filling micropores and interfacial defects, thereby promoting effective bridging and stress transfer, which enhances interfacial bonding strength and material toughness. A three-dimensional reinforcement network is established to redistribute stress, delay the damage accumulation process, and enable the formation of conductive pathways for self-sensing and damage monitoring. Multi-parameter synergistic responses, including a sudden increase in ringing counts, an initial decrease followed by an increase in the b-value, nonlinear resistivity variation, and synchronized strain, are all the key precursor characteristics of the transition from damage accumulation to instability failure. Coupled predictive analysis indicates that the b-value prediction interval is 93–96%, the RA-AF critical slowing down (CSD) variance prediction interval is 87–92%, and the autocorrelation coefficient prediction range is 85–93%. This study provides new insights into the design and damage monitoring theory of advanced lining materials for underground gas storage in abandoned mines.