Multiproperty Optimization of Steel Fiber–Reinforced Desert Sand Ultrahigh-Performance Concrete Using Response Surface Methodology and Grey Relational Analysis

Sixteen ultrahigh-performance concrete specimens containing desert sand and crushed aggregate (DS-CAUHPC) were prepared by effectively utilizing local desert sand. The mechanical and durability properties of the concrete were systematically investigated, and the results showed that steel fiber incorporation could effectively alleviate the brittleness and increase the tensile strength, and the 28-day tensile strength could reach 7.7 MPa at a desert sand replacement rate of about 40% with a crushed aggregate replacement rate of 10%. The best flowability was achieved at a desert sand replacement rate of 20% with a crushed aggregate replacement rate of 20%, which reached 681.2 mm. The best 28-day compressive strength of 142.2 MPa was obtained at 20% desert sand replacement with 30% crushed aggregate replacement. A 28-day modulus of elasticity of 45.2 GPa was achieved at 40% desert sand replacement with 20% crushed aggregate replacement. Scanning electron microscope (SEM) analysis of freeze–thaw specimens showed that the higher the desert sand content, the more obvious the formation of microstructural cracks. However, the specimens still meet the requirements for freeze–thaw resistance. The relationship between mechanical properties and loss rate was modeled using response surface methodology (R2 values of the predictive models were all greater than 0.9). In addition, a comprehensive assessment of the performance of DS-CAUHPC was conducted by combining an integrated approach of weighted analysis and grey correlation theory. This dual-method framework links theoretical metrics to practical constraints, facilitating material optimization for industrial applications.