Probabilistic Modelling and Reliability Assessment of Mechanical Properties of PSA–Cement Stabilized Lateritic Soils for Pavement Applications

This study presents a probabilistic modelling and reliability-based assessment of the mechanical properties of periwinkle shell ash (PSA)–cement stabilized lateritic soils for sustainable pavement applications. Lateritic soils, though abundant in tropical regions, often exhibit high variability and inadequate strength for engineering use without stabilization. In this research, PSA, an environmentally sustainable waste material, was utilized as a supplementary cementitious component to enhance soil performance. Response Surface Methodology (RSM) based on a Central Composite Design (CCD) was employed to develop predictive models for California Bearing Ratio (CBR), Unconfined Compressive Strength (UCS), and Indirect Tensile Strength (ITS). The developed models were integrated into a reliability framework through the formulation of limit state functions corresponding to standard engineering thresholds. Reliability indices and probabilities of failure were computed using the First-Order Reliability Method (FORM) and validated with Monte Carlo Simulation (MCS).The results indicate reliability indices ranging from 1.6192 to 1.7841, with corresponding probabilities of failure between 3.72% and 5.27%. The close agreement between FORM and MCS demonstrates the robustness and accuracy of the modelling approach. Among the evaluated properties, UCS exhibited the highest reliability, while ITS showed comparatively lower reliability, indicating susceptibility to tensile cracking. Overall, the findings confirm that PSA–cement stabilized lateritic soils are suitable for sub-base applications under low to medium traffic conditions. The study highlights the importance of reliability-based design in geotechnical engineering and supports the adoption of sustainable materials in pavement construction