Multi-Performance Evolution and Elasto-Plastic Damage Modeling of Basalt Fiber-Reinforced EPS Geopolymer Lightweight Concrete

To elucidate the multi-performance evolution mechanisms of basalt fiber-reinforced lightweight expanded polystyrene geopolymer concrete (LEGC), a two-tiered investigation was conducted. In the first part, a series of LEGC mixtures with varying volume fractions of EPS (10-40%) and basalt fiber (BF) (0.4-0.8%) were designed. Experimental tests were carried out to evaluate density, flowability, compressive strength, flexural strength, and splitting tensile strength. Crack propagation behavior was monitored using DIC-3D speckle imaging. Additionally, X-ray CT scanning revealed the internal clustering of EPS particles, porosity distribution, and crack connectivity within LEGC specimens, while SEM analysis confirmed the bridging effect of basalt fibers and the presence of dense matrix regions. These microstructural observations verified the consistency between the synergistic effects of EPS weakening and fiber reinforcement at the microscale and the macroscopic failure behavior. The results indicated that increasing EPS content led to reduced mechanical strength, whereas the reinforcing effect of basalt fiber followed a rising-then-falling trend. Among all specimens, LEGC20BF06 exhibited the best comprehensive performance, achieving a compressive strength of 40.87 MPa and a density of 1747.6 kg/m³, thus meeting the criteria for structural lightweight concrete. In the second part, based on the experimental data, predictive models were developed for splitting tensile and flexural strengths using compressive strength as a reference, as well as a dual-factor model incorporating EPS and fiber contents. Both models were validated and demonstrated high predictive accuracy. Furthermore, a splitting tensile elasto-plastic damage constitutive model was proposed based on composite mechanics and energy dissipation theory. The model showed excellent agreement with experimental stress-strain curves, with all fitting coefficients of determination (R²) exceeding 0.95. These findings offer robust theoretical support for the performance optimization of LEGC and its application in green construction and prefabricated structural systems.