Abstract ID: 3.11119 | Accepted as Poster | Poster | TBA | TBA

Debris flows pose significant geohazards to critical infrastructure, particularly water, oil, and gas pipelines in mountainous regions. Understanding the impact forces exerted by debris flows on pipelines is essential for enhancing their resilience. This study investigates the relationship between debris flow rheology—including shear stress, shear strain, viscosity, and normal stress—and the resulting impact forces on pipelines. The findings contribute to improved pipeline design, reducing failure risks and ensuring the safe operation of energy systems. In this study, experimental data were collected from debris flow samples with solid volume fractions ranging from 0.20 to 0.80 (S0–S7). The rheological properties were analyzed using a digital hybrid rotational rheometer with vane rotor and parallel plate geometry systems. The collected samples were prepared from reconstituted debris flow sediments located in Cameron Highland, Malaysia. The findings of the study indicate that increasing solid volume fraction (Cv) enhances yield stress and viscosity, exhibiting non-Newtonian behavior consistent with the Herschel-Bulkley model. The consistency coefficient (k) ranged from 0.00035 to 10.43 Pa·sⁿ, while the pseudoplastic index (n) varied from 0.16 to 1.91. While the yield stress was notably influenced by substituting 6% of coarser particles with finer material in samples S3 and S4, highlighting the critical role of particle size in debris flow mobilization. Moreover, machine learning techniques, including Random Forest and Gradient Boosting Machines (GBM), were employed to predict impact forces using experimental data. The Random Forest model demonstrated high predictive accuracy (RMSE = 0.12, R² = 0.95), while the GBM model achieved RMSE = 0.15 and R² = 0.93. These results emphasize the significant influence of rheological properties on impact forces and highlight the potential of data-driven modeling in geohazard mitigation. By integrating experimental rheological analyses with predictive modeling, this research offers valuable insights for improving pipeline resilience in debris flow-prone regions. The findings support the incorporation of rheological parameters into infrastructure design, aligning with sustainable engineering practices. Future work will focus on real-time monitoring and early warning systems to further enhance infrastructure safety.