Statistical Analysis of Tensile Damage of Basalt Fiber Foam Concrete Based on DBSCAN Clustering Method

The study proposes a methodology that combines digital image correlation (DIC) with cluster analysis (CA) to investigate the damage evolution and localization behavior of basalt fiber foam concrete (BFFC) under tensile loading. This method can simultaneously conduct quantitative analysis of both the process of damage accumulation and the process of damage localization. Quasi-static tensile tests were performed on specimens with different matrix densities and basalt fiber content. The full-field and full-process deformation images of the specimens were recorded by a high-resolution CCD. Cluster analysis was performed on the precise deformation data obtained from the DIC method, and damage extent factors and damage localization coefficients were defined. Statistical analysis indicates that the incorporation of basalt fibers not only effectively delays the progression of damage in foam concrete materials but also significantly enhances their initial damage threshold load and inhibits the phenomenon of damage localization in foam concrete. Compared to specimens without basalt fibers, those incorporating basalt fibers exhibited increases in the damage localization coefficients at tensile failure of 0.4, 0.33 and 0.18, respectively, under three different matrix density conditions. Therefore, the proposed DIC-CA method, in conjunction with the defined damage extent factor and damage localization coefficient, can effectively and quantitatively capture the two key dimensions of damage (accumulation extent and spatial distribution characteristics) in fiber-reinforced foam concrete under tensile loading. This provides an efficient, intuitive, quantitative analysis method for characterizing the initiation, development and localization processes of damage in similar materials.