Project 2.2: Development of computational methods for risk assessment of dams using high performance computing

Dams are complex systems exposed to multiple hazards, including earthquakes, floods, material degradation, and human-induced risks. Traditional risk assessment approaches often treat these hazards independently, overlooking their interactions and compounding effects. This work proposes a unified probabilistic framework for multi-hazard risk analysis of dams, integrating both physics-based and data-driven methodologies. Physics-based models ensure accurate representation of structural mechanics under static and dynamic loads, while data-driven approaches leverage advances in sensing technologies and machine learning to assimilate field measurements. Together, these approaches enable the development of digital twins of dams, offering predictive insights for maintenance and safety management. Key research directions include scalable computational methods for high-performance computing, domain decomposition for static analyses, reduced-order modeling for dynamic risk prediction, and non-intrusive techniques compatible with black-box solvers. Physics-informed machine learning will be employed to bridge mechanistic understanding with real-time data, enhancing predictive accuracy. The probabilistic framework will incorporate Monte Carlo simulations, spectral stochastic finite element methods, reduced-order models, and neural networks to quantify uncertainties and propagate them through computational models. Outcomes will include validated computational tools, machine learning models for risk prediction, and guidelines for multi-hazard risk assessment tailored to Indian dams. By combining advanced simulations with field data, the project aims to provide actionable strategies for dam safety, predictive maintenance, and resilience under evolving climatic and operational conditions. The contributions will strengthen dam risk management practices, reduce vulnerability of downstream communities, and advance the integration of digital technologies in infrastructure safety.