Analysis of Variables Influencing Scour on Large Sand-Bed Rivers Conducted Using Field Data

Throughout the lifespan of a bridge, morphological changes in riverbed affect variable action-imposed loads on structure. This emphasizes need for accurate and reliable data that can be used model-based projections targeted identification risk associated with bridge failure induced by scour. The aim this paper is to provide an analysis scour depth estimation large sand-bed rivers under clear water regime, detect most influential (i.e., explanatory) variables, examine relationship between them as response variable. A dataset was obtained from United States Geological Survey’s extensive field database local at piers, i.e., Pier-Scour Database (PSDB-2014). original filtered exclude did not reflect rivers, several variables were omitted using principal component analysis. reduction process resulted 10 multiple non-linear regression modeling (MNLR). Two MNLR models non-dimensional dimensional models) prepared estimation; however, model slightly overperformed other one. According Pearson correlation coefficients (r), estimating follows: Effective pier width (r = 0.625), flow 0.492), critical velocity 0.474 r 0.436), respectively. In compounded hydraulic-sediment category, had greatest impact highest coefficient) comparison densimetric Froude numbers characterized 0.427 0.323, remaining four (local bed shear stress, number, particle Reynolds number) exhibited very weak depth, < 0.3.