ABSTRACT

Revetments built on concave banks in bends of sand-bed rivers formed in alluvium can be undermined as the streambed scours. For this reason, protection work along the banks needs to extend to the lowest level reached along the toe of the underwater slope. To obtain robust, easy to apply, relations for rapidly predicting the maximum depth of bend scour in alluvial rivers where the streambeds are composed primarily of sand-sized sediment, on-site measurements made by several investigators are assembled and used to fit coefficients of an artificial neural network model. Model variance is also calculated, which provides a means of generating the prediction intervals of bend-scour depths. Consequently, one can compute a design scour-depth margin of safety that is statistically well-reasoned and not overly excessive.