To determine the cause of the collapse of the spans of the u.S. 51 Bridge over the hatchie river in 1989, a 1:220 scale model of bent 70 (failure of this triggered the collapse) was tested. The testswere designed to determine how the maximum local pier scour may have ocurred after the channel migrated to bent 70, and to obtain videotape shots of the local scour process for use as a visual aid in thepublic hearing conducted to gather evidence concerning the collapse. The article describes the tests, the scaling ratios, the experiments, and the results. It was found that local pier scour was not the primary element in the collapse, it probably was a contributing factor for a foundation that was already vulnerable once the channel migrated laterally. Local pier scour from the complex foundation was nomore than it was from the simpler and smaller squre column by itself. It is recommended that for bridges such as this one which have channel migration and unknown local scour during floods, a convenient instrument be developed to facilitate recording stream bed cross sections just upstream and downstream of the bridge from the bridge deck.
Abstract