Modeling Shipping Container Velocity during Tsunami Conditions

by Alexis Mills, Humboldt State University,
Bryan Peele, North Carolina State University,
Krystian Paczkowski, University of Hawaii at Manoa

The risk that tsunamis pose to coastal structures is not just from the wave impact, but also from the impact of debris that is carried by the fluid flow. Shipping containers are ubiquitous in coastal locations and represent a realistic debris object. For want of a better approximation, it is often assumed that the debris speed is equal to the fluid speed.

To understand better how a tsunami bore can accelerate shipping containers, ANSYS Fluent Computational Fluid Dynamics (CFD) flow modeling software was used to calculate tsunami-induced velocity profiles for floating shipping containers. Simulations were used to estimate the rate at which the container velocity approaches the on-shore flow velocity of the bore, which can approach 20 miles per hour or greater. At its maximum weight, a 40-foot-long sealed shipping container weighs more than 60,000 pounds and will float with a draft of 0.96 meters.

2-D cut of mesh around the container
Figure 1: 2-D cut of structured region following the container motion as the surrounding unstructured region is dynamically meshed around it.

For the simulations, the primary objective was to determine the velocity of the container, which was calculated as an average velocity along the flow field at the container boundary. For this reason, it was necessary that the computational model had a refined mesh for the velocity field surrounding the container and for this cell refinement to follow the path of the container as it moved through the domain. The simulations used a six degree of freedom rigid body model for the shipping container as it moved through the domain. ANSYS Fluent requires a structured mesh directly surrounding the moving container with the remaining domain unstructured.

Pointwise allowed the straightforward creation of these hybrid mesh topologies. Initial tests were run using 2D meshes as a verification process for the computational model, creating a reference for a 3D model and verifying that the 3D model would produce meaningful results.

Because various scenarios with changing container placement and inlet conditions were tested, it was critical that Pointwise allow quick and easy alterations in geometry. These alterations in geometry were aided by the efficient layering system incorporated into Pointwise. This system allowed the geometry and mesh components to be easily organized for quick modification and visualization. Large sections of the geometry may be translated at once and Pointwise will automatically redefine the mesh to adapt to these changes. With many other software packages, it would be necessary to redefine each block separately, a much more tedious and time consuming process.

Pointwise operates in a way that allows easy mesh integration with ANSYS Fluent. Boundary conditions assigned in Pointwise are always recognized by Fluent, saving time and effort. Unlike many other meshing software, there was no unnecessary separation into multiple boundary elements when imported into Fluent.

Mesh colored by cell skewness
Figure 2: Pointwise offers many examination techniques, including coloring cells based on cell skewness.

The ability within Pointwise to view cross-sections of a 3D mesh offers unparalleled access to visualize detailed mesh interiors. Pointwise offers many examination techniques including coloring cells based on cell volume, skewness and aspect ratio (Figure 2). Since cell shape and volume are important factors in the time-step dependent dynamic meshing, it was vital that the meshing software be capable of displaying these attributes. Without proper cell arrangement and sizing, the model will provide unrealistic results. Monitoring these attributes throughout the meshing process ensures a quality mesh as the final product. Similar mesh monitoring tools would be beneficial in post-processing if meshes could be imported from ANSYS Fluent after simulation when the mesh has dynamically transformed.

The use of Pointwise created a seamless interface between the meshing process and CFD simulations. Pointwise saved significant time on this project by allowing simplified grid generation and redefinition, as well as extended mesh monitoring capability. Knowledge gained from this project will be incorporated into larger studies aimed at better designing coastal structures to withstand debris impact. Computing shipping container acceleration during tsunamis with respect to time and distance propelled will also assist in quantifying risk to port facilities.

This work was carried out under a National Science Foundation Research Experiences for Undergraduates (REU) summer program, site grant #0852082, at the University of Hawaii. The first two authors gratefully acknowledge the financial support provided by this grant. The mentors for this program were Krystian Paczkowski, Ph.D . student, and Professor H.R. Riggs.

Search Articles

Share This