Research Projects

India ranks third globally in the number of large dams, with over 5,200 operational and nearly 460 under construction (ICOLD, 2017). These dams play a critical role in water management, irrigation, energy production, and flood control. Ensuring their safety and functionality is vital for national water security. Many dams in India are aging, with nearly 200 large dams now over 100 years old. These aging dams require a systematic approach to rehabilitation, which includes detailed condition assessment, proper design and execution of repair works, and strict quality control measures. The failure of a dam can lead to devastating consequences, including loss of life, damage to infrastructure, and major economic setbacks. Literature indicates that seepage-induced instabilities have caused almost 50% of embankment dam failures globally (ICOLD, 2015). In India, most reported dam failures between 1951 and 1970 involved embankment dams. This highlights an urgent need to better understand seepage processes and their role in dam instability, particularly for earthen dams. Traditionally, seepage analyses have been carried out using continuum methods, such as the Finite Element Method (FEM), based on macro-scale parameters. While these approaches are useful for engineering applications, they may not fully capture the fundamental particle-scale mechanisms—such as internal erosion—that often initiate dam failure. Some recent efforts have attempted to couple Computational Fluid Dynamics (CFD) with the Discrete Element Method (DEM) to address seepage and slope instability problems. However, DEM alone becomes computationally expensive for large-scale problems like embankment dams. This project proposes a multiscale modelling approach that integrates FEM with DEM to address this limitation. In this framework, FEM will handle the overall boundary value problem (e.g., the full embankment), while DEM will be used to simulate small-scale Representative Volume Elements (RVEs) at the Gauss points in the FEM mesh. These RVEs will provide realistic material responses based on particle interactions, bypassing the need for complex constitutive models. The main scientific objective is to investigate how particle-scale processes influence the macro-scale behaviour and stability of embankment dams. The hypothesis is that a multiscale FEM-DEM approach can better capture the initiation and progression of internal erosion and provide a more accurate assessment of dam safety. The outcomes of this research will deepen our fundamental understanding of seepage-induced failures and support the development of more reliable design and rehabilitation strategies for hydraulic structures. This project will have direct relevance to dam safety programs in India, such as the Dam Rehabilitation and Improvement Project (DRIP), offering practical insights for risk mitigation and asset management.