Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1 CHAPTER 1: INTRODUCTION Current application of scour prediction methods overestimates scour depths around abutments and in contracted openings at many locations. Such excessive scour depth prediction results in construction of unnecessarily deep foundations or installation of unnecessary countermeasures. State Departments of Transportation are using methods recommended in HEC- 18 (Richardson and Davis, 2001) to estimate the potential for scour at bridges. These scour- prediction methods are based primarily on simplified small-scale model studies conducted in laboratory flumes. These laboratory investigations typically model straight, rectangular channels with uniform approach-flow velocities, approach-flow depths, and non-cohesive bed material. The floodplains represented in the model studies are often of uniform roughness and are typically of a roughness similar to the main channel; however, variable width compound channels, floodplains with highly non-uniform roughness and non-uniform sediments with varying degrees of cohesiveness, are typical of most bridge sites. The concept of scour components, which can be summed to obtain total scour, was derived from observations of scour in laboratory studies. These components consist of local scour at piers and abutments and contraction scour caused by the overall contraction of flow at the bridge. Long-term degradation associated with the streambed adjustments over long river reaches, considered independent of the bridge, also contributes to changes in streambed elevations at the bridge; therefore, long-term degradation is considered as a component of total scour.
2 Error in prediction of scour components stems from three sources: (1) estimation of hydraulic parameters, typically through hydraulic modeling; (2) selection of scour-prediction parameters; and (3) scour-prediction equations. The hydraulic parameters usually are estimated from a one-dimensional hydraulic model that distributes flow across the approach and bridge opening by conveyance (combination of roughness and flow area); however, the flow distribution at a bridge or in its approach is typically non-uniform because of cross-stream flow caused by channel bends, complex roughness patterns, irregular valley topography, and obstructions in the floodplain. Bridges and approach embankments not aligned perpendicular to the approach flow further complicate flow patterns and velocity distributions. The empirical scour-depth prediction equations developed from laboratory studies use average flow parameters such as approach velocity, flow depth, and embankment length. A high degree of subjectivity is often required to select these parameters. The simplifications involved in using laboratory experiments to develop scour-prediction methods and the subjectivity required to extract average representative parameters from non-uniform and heterogeneous field conditions contribute to the uncertainty and error of scour-depth prediction. A well-recognized source of scour-prediction error is the inadequate representation of erosion resistance of soils. The scour-prediction equations recommended in HEC-18 were developed for uniform, unstratified, non-cohesive sediments that are representative of the most severe scour conditions. The erosional resistance of typical soils found at bridge sites is a combination of stratified soils with varying degrees of cohesiveness. In addition, the surface soils often are protected and reinforced by vegetation or armored by the largest size fractions of the
3 bed material. The complexity of the erosion resistance of bed material has been marginally included into scour-prediction equations. Three comparisons are necessary to evaluate the current design guidance and to form the basis of significant improvement in scour-prediction accuracy. First, comparison of scour depth predicted by the current guidance with field measured scour depth is needed to provide an overall assessment of the state-of-practice. Second, comparison of the hydraulics from one- dimensional numerical models with the measured hydraulics is required to evaluate the adequacy of those models for estimating the hydraulics at contracted sites. Third, comparison of scour computed using measured hydraulics with the observed depth of scour is needed to provide a direct evaluation of the scour-prediction equations. These comparisons are the basis for determining the source of inaccuracies associated with the scour-prediction methods. PURPOSE AND OBJECTIVES The main purpose of NCHRP Project No. 24-14 was to collect field data from which processes affecting scour magnitude in contracted openings could be identified, to support verification of physical and numerical model studies, and to improve guidelines for applying scour-prediction methods at contracted bridge sites. The specific objectives of this research are as follows:
4 (1) to describe and quantify the influence of processes affecting scour magnitude in contracted openings using field data; (2) to provide field data for use in verification of physical- and numerical-model studies; (3) to develop interim guidelines for applying scour-prediction methodology at contracted bridge sites for a wide range of common field situations; and (4) to provide recommendations for future research that will advance scour- prediction methodology in accordance with the Strategic Plan for Scour Research as modified by the findings of this research (Parola et al, 1996). SCOPE AND APPROACH OF RESEARCH The objectives were accomplished by the collection and analysis of data at 15 bridge sites. A combination of real-time and post-flood data-collection activities provided comprehensive field data sets. Real-time measurements are measurements of flow velocities and channel bathymetry during the flood event. Post-flood data collection consists of detailed bathymetric, geotechnical, and geomorphologic measurements obtained after the floodwaters recede. Emphasis was placed on collection of comprehensive real-time and post-flood data sets to quantify the non-uniform and time-dependent flow and geotechnical conditions at the sites and to define the processes responsible for total scour. Scour that forms within bridge contractions and around bridge abutments is dependent on the entire flow field approaching, within and exiting the bridge area; therefore, detailed directional velocity data were collected throughout the
5 reach affected by the bridge where flood and site conditions permitted (4 of 15 sites). In addition, streambed, stream bank, and floodplain-material properties were described. Raw data were reduced and assembled into a database. The database interface was developed such that the information is easily accessible by both researchers and highway engineers. The database is accessible through the World Wide Web at the following location: http://ky.water.usgs.gov/Bridge_Scour/BSDMS/index.htm. The scour predictions based on the methods provided in HEC-18 were compared to the observed scour at each site. Flow velocity and depth data obtained from real-time investigations along with post-flood topographic surveys were used to develop and calibrate two-dimensional models such as RMA-2 and FESWMS at two sites. The one-dimensional hydraulic models HEC- RAS or WSPRO were developed for all sites where sufficient cross-sectional data were collected or available. The velocities obtained from numerical simulations were compared to measured velocities. The research team identified processes that substantially affect scour, but are not represented in HEC-18. The observations and measured data demonstrate inaccuracies of the current scour-prediction methods as specified in HEC-18; however, there were insufficient data to support the reliability of recommended changes without additional research.
6 The research team provided recommendations for future research that will advance scour prediction methods. These recommendations include suggested modifications to the Strategic Plan for Scour Research to reflect the findings of this research.
