Novel Eulerian Two-phase Simulations for Burial Dynamics of Munitions

Dr. Tian-Jian Hsu | University of Delaware

MR20-1478 Phase I


The project team presents major results from a one-year SEED project to prepare a high fidelity two-phase numerical modeling framework SedFoam, for simulating scour burial in fluidization (sheet flow) condition, interaction with bedforms, and eventually investigating the effect of munition density. This proof-of-concept research effort was led by the objectives to (1) demonstrate that the Eulerian two-phase model SedFoam provides an effective modeling framework to directly resolve critical processes in scour burial, which cannot be easily achieved by conventional single-phase modeling approach, and (2) implement a few model enhancements for more physical-based closures in SedFoam to simulate processes that may lead to deep burial, namely, fluidization and bedforms.

Back to Top

Technical Approach

SedFoam is an open-source Eulerian two-phase model for sediment transport applications. By solving the mass and momentum equations for the water phase and fluid phase with closure on turbulence, particle stress and interphase momentum coupling, SedFoam resolves the full dynamics of sediment transport without the need to artificially separate transport into bedload and suspended load layers. Although SedFoam has been applied extensively to simulate current and wave driven sheet flow transport, plug flows, three-dimensional (3D) scour around a cylindrical pile, these sediment transport problems are limited to the top 10 centimeters of the seabed and its application for deep burial processes, such as liquefaction, requires additional model enhancements. As a first step, the project team made important progress to include dilatancy effects into SedFoam. Improved SedFoam can reproduce slow and rapid movement of an underwater avalanche due to negative and positive pore-pressure response caused by dilation and contraction of sediments resulting from different initial packing concentration observed in the laboratory experiment.

Back to Top

Interim Results

Major efforts were also devoted to proof-of-concept simulations. The two-phase model SedFoam was able to simulate piping as the onset of a two-dimensional pipeline scour driven by waves similar to laboratory observation. Piping is a realistic scour onset process triggered by pore water flow passing underneath the structure because of the upstream-downstream pressure difference. This was an important demonstration of SedFoam’s capability as typical single-phase computational fluid dynamics models for scour could not allow the submerged object to be directly attached to or buried in the sand bed. Next, SedFoam was applied to simulate scour of a 3D horizontal short cylinder from low to high Keulegan-Carpenter (KC) number as a key step toward future simulation of unexploded ordnance (UXO) since horizontal short cylinder can be considered as the simplest UXO configuration having all the essential 3D characteristics. In the high KC number simulation similar to those cases observed in the field at Duck, NC (USA) when deep burial occurs, incoming boundary layer flow separation and horseshoe vortex are generated at the front of the cylinder. The vortices further extend to the two lateral faces of the cylinder as flow accelerates and leads to more significant erosion. Thirdly, the project team investigated SedFoam’s capability to simulate migration and evolution of bedforms driven by waves. SedFoam was able to predict the evolution of ripple length and height subject to changing waves with good agreements with empirical ripple predictors. Moreover, the model was able to predict onshore ripple migration driven by onshore-velocity skewed waves similar to measure data.

Back to Top


In summary, the project team delivered a validated SedFoam model capable of accurately and efficiently simulating several key coupling mechanisms between hydrodynamics, sediment transport, underwater objects and seabed response. A more extensive research effort is needed to integrate these aforementioned key components consistently in the SedFoam modeling framework to create a new numerical modeling tool for scour burial of UXO. The new model can be used as a reliable analytical tool to study the mechanisms and thresholds for deep burial. Moreover, well-designed and extensive numerical simulations can be carried out to expand and to fill the data gaps of scour burial data in the Underwater Munition Expert System.

Back to Top

Points of Contact

Principal Investigator

Dr. Tian-Jian Hsu

University of Delaware

Phone: 857-204-0336

Program Manager

Munitions Response