Modeling of Submarine Landslide Generated Tsunamis and Its Application in Risk Mitigation and Hazard Management at Nuclear Reactor Sites, 20-R9725

Printer Friendly Version

Principal Investigators
John Stamatakos
Kaushik Das
Debashis Basu
Ron Janetzke
Steve Green

Inclusive Dates:  01/01/07 – 04/31/09

Background - Tsunamis and associated flooding are natural hazards that pose potential risks to critical facilities located near coasts. Tsunamis generated by large-magnitude earthquakes in the sea floor are relatively well understood. However, tsunamis also result from large submarine landslides, and their effects are less well known. Additional research is needed to develop a more comprehensive understanding of these waves as hazards. Though significant progress has been made in understanding the entire tsunami evolution process from wave generation to wave runup, a more realistic and validated simulation methodology needs to be developed for improved prediction of submarine landslide generated tsunamis. Most of the tsunami models to date rely on shallow water equations.

Approach - Advances in computational fluid dynamics and parallel computing have led to the application of full three-dimensional Navier-Stokes equations for tsunami simulations. However, modeling of landslide-generated tsunamis using the full Navier-Stokes equations presents a significant computational challenge because of the complex motions and shapes of the waves. These complexities, along with the three-dimensionality of the flow field, require alternative methods that can simulate these complex flows with greater accuracy and relatively fewer computational resources. As an alternative, recent advances in mesh-free particle-based computing methods such as smoothed particle hydrodynamics and discrete element methods may provide equally valid models of landslide-driven tsunamis, which are considered in the present research along with a traditional Navier-Stokes equation-based flow solution approach. The emphasis of the investigations was to develop an integrated computational methodology using traditional grid-based Eulerian techniques and mesh free particle method to evaluate tsunami hazards at coastal nuclear installations and structures caused by submarine landslide induced fluid motion such as tsunamis.

Accomplishments - The successful completion of the project resulted in the development of a methodology for simulation of landslide-generated tsunamis using both the Navier-Stokes volume of fluid (VOF) method and smoothed particle hydrodynamics (SPH) method. Simulations carried out using the Navier-Stokes simulations and commercial software FLOW-3D highlighted both the importance and complexity of slide deformation, three-dimensionality, slide geometry and surface roughness height on wave characteristics, and the importance of slide rheology in tsunami risk assessment. Simulations were carried out for the Lituya Bay landslide generated tsunami in Alaska, and the simulations were compared to the available scaled-down experimental data. The computed results had a reasonably good match with available experimental data. A new multiscale turbulence model has been implemented in FLOW-3D to study the effect of turbulence modeling and turbulence diffusion on the vortex structure in the tsunami waves. These investigations and developed computational tools emphasized the importance of slide geometry, deformation, and viscosity as well as boundary conditions on the predicted wave characteristics and run-up height. Four full-length conference papers resulted from this research, and three papers have been submitted to peer-review journals for publication. In addition, project staff members have prepared promotional material based on the research and are actively developing business and collaboration opportunities.

2009 Program Home