Experimental Study of Tsunami-Induced Loads on Bridges

Abstract

New Zealand is subject to considerable local tectonic activities and lies exposed to major earthquake events in the Pacific Ocean, which makes it quite vulnerable to tsunami events. Over 14 major tsunami events struck the world’s coastlines, causing death and economic destruction, between 1990 and 2000 (Bryant, 2014), and at least 32 tsunami have been recorded in New Zealand since 1840 (De Lange et al., 1986). The disastrous Indian Ocean Tsunami of December 26, 2004, and the Japanese Tohoku Tsunami of March 11, 2011, highlighted the weaknesses and defects in preparation and warning systems (Bryant, 2014). A perusal of the available guidance for estimating tsunami loads on coastal structures and bridges shows a large variation in the values estimated by the proposed equations and the lack of general agreement even on the types of forces imposed on the bridge decks. Reviewing the available literature revealed that the magnitudes of the actual tsunami loads on coastal structures and bridges are usually not accurately known. It emphasises the need for more investigation of tsunami interaction with coastal structures, as that could be of benefit for designing bridges that may be subjected to tsunami loads. The main aim of the present study is to experimentally examine the interaction of a tsunami bore with coastal bridges. To accomplish this objective, physical modelling of the tsunami bore in the laboratory was conducted by investigation of (1) the impact of a tsunami bore on a box section bridge deck with different deck clearances; (2) the impact of a tsunami bore on a deck-girder section bridge with different deck clearances; (3) the effect of contraction on the tsunami induced pressures and forces on a box section bridge deck with wing wall and spill-through abutments with different lengths and deck clearances; (4) the impact of a tsunami bore on a skewed box section bridge deck with different skew angles and deck clearances; (5) the performance of a skewed deck-girder section bridge under tsunami bore impact with different skew angles, deck clearances and handrails (i.e. solid and porous handrails). The experiments were performed in Flumes A and B. The experiments to investigate the effect of contraction on the tsunami induced pressures and forces on a box section bridge deck with wing wall and spill-through abutments were undertaken in Flume A, which had a 14 m long, 1.2 m wide and 0.8 m deep wave flume, connected to a reservoir 11 m long, 7.3 m wide and 0.6 m deep. In order to investigate the interaction of a tsunami bore with bridges in a broader range of bore height and velocity, a tsunami wave flume (Flume B) at the newly commissioned laboratory facilities at the University of Auckland was constructed and utilised. All the remaining test cases were conducted in Flume B (i.e., unskewed and skewed box section and deck-girder section bridges), which is connected to a 6.4 m long, 5.5 m wide and 1.2 m deep reservoir made of concrete block walls. It should be noted that both flumes were separated from the reservoir using an automatic gate, which is designed to be lifted to a constant height to generate a tsunami bore with different bore heights and velocities. Capacitance-type wave gauges were placed at the centre of the flume along its length to measure the bore depth and velocity. A multi-axis waterproof load cell was used to measure the applied forces and moments, and pressure sensors were used to measure the horizontal pressures applied on the bridge deck. For the impact of a tsunami bore on a box section bridge deck, a simplified box section bridge made from acrylic sheets was used with different deck clearances. Pressure transducers were also used to measure the tsunami induced pressures on the bridge deck. A good agreement between the forces measured by the load cell and those derived from the pressure transducers and the forces theoretically computed with the equations available in the literature showed the validity of the experimental results. The horizontal force was theoretically calculated based on the assumption that the total horizontal force on a bridge deck is mainly correlated with tsunami bore flow velocity (and to a lesser extent with tsunami bore height) and is equal to the sum of the hydrodynamic and hydrostatic forces. The forces and moments induced by the tsunami bore and acting on a box section bridge deck with wing wall and spill-through abutments, were measured. The applied forces and moments were found to be larger for the bridge with spill-through abutments compared with the bridge with wing-wall abutments. For the impact of a tsunami bore on a skewed box section bridge deck, lateral force (Fy) and rolling (Mx) and yawing moments (Mz) were introduced, which had been zero for the unskewed bridge deck. The results showed a decreasing trend for the horizontal and vertical tsunami forces and pitching moments when increasing the skew angle of the bridge deck. In contrast, an increasing trend was observed for the lateral force and rolling and yawing moments when increasing the skew angle. Similar to the skewed box section deck, for the impact of a tsunami bore on a skewed deck-girder section, the time histories of the induced forces and moments showed a similar shape. The main difference is that the normalised horizontal forces on the box section deck are higher than those on the deck-girder section bridge. In contrast, the normalised uplift forces on a deck-girder section were found to be higher than those applied on a box section bridge deck. The bridge girders trapped the air and water, which acted as a damper and reduced the horizontal forces for the girder bridge deck. Unlike the horizontal forces, the trapped water and air between the bridge girders increased the buoyancy force and subsequently the uplift forces on the bridge deck. Lateral force and rolling and yawing moments were the additional components observed due to the different impact mechanism of the bore front on a skewed deck. The asymmetrical geometry of the skewed bridge deck on both sides of the vertical plane of symmetry caused water diversion towards the flume sidewalls in the Y direction, which exerted a lateral force on the deck. The front-acute corner of the bridge deck was the first point of the deck hit by the bore front, which lifted one side of the deck and rotated it about the flow direction (Mx). The flow then spread over the entire span. This lag in the period of the initial impact introduced the yawing moment (Mz). Mz rotated the bridge in the horizontal plane. Both Mx and Mz were negligible for the unskewed bridge. The outcome of this study would be a preliminary guideline for tsunami loading characteristics that could be of benefit for researchers and design engineers to have a better understanding of tsunami bore impact on coastal bridges. Based on the experimental results, equations proposed for estimating tsunami horizontal, lateral and uplift forces for a box section bridge deck and a deck-girder section bridge including the effects of contraction and skewness. The available equations in the literature were adopted on an “average” New Zealand road bridge to quantify the tsunami induced loads on it under different tsunami scenarios. Also, a comparison of the equations from the currently available guidance was made for estimating tsunami loads on an “average” NZ road bridge to calculate the most suitable equations for design purposes.