Evaluating and Understanding of Bridge Scour Calculation

Abstract

Channel scour and channel instability at bridges are the major cause of bridge collapse caused by flooding. Currently, Departments of Transportation (DOTs) in the USA use the HEC-18 procedure and associated software, which was developed for the scour depth calculation of non-cohesive soil, for scour calculation of any soil. Scour in non-cohesive soil is well known but scour in the cohesive soil is not. HEC-18 provides a deterministic procedure to calculate the scour depth near a bridge site, but the procedure and the input parameter determination have various uncertainties, calculated scour depth could be quite different in some cases. This type of uncertainty of HEC-18, i.e., the variability of both the input parameters and calculated scour depth, is dealt with in this study. The critical shear stress for scour in HEC-18 is linked with median particle-size (D50) but actually related to many other factors especially for cohesive soils. HEC-18 critical shear stress and critical velocities of six clay soil samples are compared to the critical shear stress and critical velocity previously obtained from Erosion Function Apparatus (EFA) for studying the uncertainty of HEC-18 due to the soil property. The multilayer method was proposed and tested to determine the total feasible scour depth calculated using layer by layer D50 along the depth. Comparing with the scour depth determined using average D50, which is the current normal practice by DOTs, the multilayer method can give a more reasonable prediction of scour depth but requires D50 values below the estimated scour depth. HEC-18 lacks clear instructions in determining hydraulic parameters for the scour calculation in different parts of the channel cross sections. Various one-dimensional models such as WSPRO and HEC-RAS can be used to determine the hydraulics of bridges. Although these models use the standard step method to solve the energy equation for gradually varied flow, these models have their own processes to solve the energy equation that are different from each other in many ways. The HEC-RAS models for four bridge sites in Alabama were developed by using the input data of the WSPRO models to calculate the hydraulic parameters needed for HEC-18 to calculate the scour depth. The differences in the hydraulic parameters and eventual scour depths were discussed and analyzed for understanding and evaluating the uncertainty of hydraulic parameters. A step by step procedure was developed using Pugger mixer to mix the different soil components and create non-slaking soil samples for EFA testing in order to study critical shear velocity and erosion rate of cohesive soils in the future. This study confirmed that there exists certain uncertainty in HEC-18 which should be addressed, and it is better suited for calculating scour depth for non-cohesive soil. The uncertainty of predicting and estimating the scour depth comes from various sources such as soil properties (D50, critical velocity and scour rate) and hydraulic parameters. EFA testing results could help to reduce uncertainty of scour calculations.