Halldor Arnason1, Harry Yeh2, and Catherine Petroff3
1 VST Engineering, Iceland
2 Oregon State University
3 LP4 Associates
We report laboratory results and data from physical modeling of free-standing structures of simple shapes, subjected to a passing of a tsunami modeled by a turbulent bore. Measurements were made of water-surface levels and velocities of the passing water as well as the forces acting on each structure. The goal is to further our understanding of the interaction between the bore and the structure, i.e. the structure’s effect on the bore as well as the bore’s effect on the structure, including the prediction of hydrodynamic force on a structure during impact and passage of a tsunami. This work was supported by the U.S. National Science Foundation (CMS-9614120, 0245206 and SBE-0527699).
Experimental Facilities and Procedures
Experiments were performed in a 16.6 m long, 0.6 m wide and 0.45 m deep wave tank at the University of Washington ( Fig. 1). A single bore is generated by lifting the 6.4 mm thick stainless steel gate, which initially separates the thin layer of water (20 mm) from the impoundment behind the gate. The gate is lifted by a 64 mm diameter pneumatic piston driven by 0.5 MPa air pressure, which results the gate to reach the top of the tank in 0.2 s. This bore generation scheme enables us to achieve the experiments precisely repeatable: the similar generation scheme was used previously by Yeh, et al. (1989) and Ramsden (1993).
Figure 1. Schematics of the experimental apparatus.
The bottom of the tank is made of stainless steel, except for a 1.5 m long transparent plexiglas bottom placed 5 m downstream of the gate. An obstacle was placed over this transparent bottom so that the flows around the column are measured optically from the underneath. The obstacle used in the experiments are a cylindrical shaped column 140 mm in diameter and a square shaped 120 × 120 mm column; the both columns are made of 6.35 mm thick acrylic to ensure their sufficient rigidity. The column is supported from the above so that all the hydrodynamic loads be effectively transferred to and accurately measured by the load cell (a multi-component force transducer). The load cell is fastened inside the acrylic column and is then connected to a rigid vertical aluminum channel (6.35 mm thick). The vertical aluminum channel is connected to a thick horizontal aluminum beam (9.53 mm thick) that lies across the tank and rests on instrument rails mounted on top of the tank’s sides (see Figure 2). The load cell is calibrated in situ using a digital load gauge.
To obtain a temporal variation of the spatial water-surface profile, the laser induced fluorescent technique was applied. A two-dimensional vertical plane was illuminated from the above by a 4-W Argon-ion laser sheet. The resonant scanner was used to convert the laser beam to the sheet of the uniform thickness. The illuminated intersection of the laser sheet with the water surface identifies the water surface profile. This technique yields temporal and spatial variations of the water-surface profiles accurately in a non-intrusive manner. Detailed procedure can be found in Gardarsson (1997) who applied this technique to investigate the water sloshing motions in tuned liquid dampers.
Figure 2. Views of load cell installation.
Flow velocities were measured with two independent instruments: 1) a digital particle image velocimeter (DPIV) and 2) a two-component back-scattering Laser Doppler Velocimeter (LDV). The DPIV measures the flow velocity in a horizontal two-dimensional plane illuminated by a pulsating dual-head Continuum Nd:YAg laser sheet. To capture the motion of the seeded reflective particles (used as the flow tracer) in the laser illuminated horizontal plane, the high-resolution CCD PIV camera was placed underneath the transparent tank floor shooting vertically upward. The sampling rate of the DPIV was 14 Hz. The DPIV was used to measure flow fields by repeating the experiments. While the DPIV can capture the flow field, the LDV can only measure velocity at a single point at a time, but the sampling rate is much higher. It measures horizontal two-dimensional velocity components by shooting the laser beams from the underneath of the transparent tank floor. Both DPIV and LDV measure the horizontal velocity components; the DPIV provides the spatial velocity field at the slow sampling rate, while the LDV provides a point measurement at the fast rate. Those two measurement techniques complement each other to yield the better understandings of the flow structures.
To synchronize all the measurement devices, the time-stamp trigger system was setup so that when the gate was released, a precise clock started for the visual data as well as sending a signal to the data acquisition system for the load cell, the wave gage, and the DPIV and LDV. In this study, the time origin was set zero based on this time-stamp trigger, i.e. at the moment of gate release. A detailed discussion of the laboratory apparatus can be found in Arnason (2005).