Hydraulic Modeling of Bridges in North Carolina

Green Street bridge over the Tar River

Background

The North Carolina Department of Transportation (NCDOT) Hydraulics Unit is responsible for providing designs for bridges, culverts, and other drainage control features while adhering to permit requirements that help maintain wetlands, habitat, and water quality. The Unit also is responsible for providing field data and hydraulic design recommendations to the Bridge Maintenance Unit for its ongoing bridge-replacement program.

In order to provide hydraulic design recommendations for the construction of new bridges and/or embankments, the NCDOT relies on the application of numerical models to simulate bridge hydraulics. The Hydraulics Unit almost exclusively uses one-dimensional (1-D) step-backwater models such as HEC-RAS, HEC-2 and WSPRO to simulate the hydraulics for bridge design regardless of the site conditions. Many bridges, however have site-specific characteristics that produce lateral and/or vertical variability in the velocity distribution that cannot be accurately represented with a 1-D step-backwater model.

Site characteristics, such as upstream channel alignment (meanders), geomorphic setting of the river valley, flood-plain alignment relative to the channel, and hydraulic structures (levees, dikes, etc.) can produce hydraulic complexities at a bridge that cannot be simulated adequately by a 1-D model. For example, channel alignment can present a unique, site-specific problem at bridges because approaching flow distributions at flood stages can be significantly altered when flows leave a channel and enter a flood plain at a channel bend. When channel bends occur just upstream from a bridge, concentrated channel flows can be directed to a section of a bridge opening that would not typically experience this magnitude of flow if the channel were straight. Under these conditions, the flow fields are inherently two-dimensional (2-D) and cannot be accurately represented with a 1-D model. The ability to accurately simulate velocity magnitudes and distributions at bridge crossings is essential in assuring that bridge plans balance public safety with the most cost-effective design.

Objectives

The objectives of the initial phase of the study are to

• provide the NCDOT hydraulics unit a 2-D numerical modeling workshop to train staff engineers in the process and techniques associated with 2-D modeling at bridges;
• develop a 1-D step-backwater model and 2-D hydrodynamic model for the US-13 bridge over the Tar River at Greenville and compare results to field data to evaluate the ability of each model to represent field conditions. This information will help to provide an initial basis for ongoing development of modeling guidelines that will ensure cost-effective hydraulic analysis.

Approach

The U.S. Geological Survey (USGS) will first provide a multiday workshop on 2-D modeling at bridges for selected NCDOT staff members. The workshop will include information on the techniques, procedures, and data required for 2-D modeling.

The main component of the study is the development and comparison of the 1-D and 2-D models. For the first phase of the ongoing study, these models will be developed for the Hwy-13 bridge over the Tar River, a complex riverine bridge site selected by the NCDOT and USGS. The criterion for selecting this and future bridge sites is to target locations with historical flood data (channel profile, discharge, velocity distribution, and stage), which could supplement or be used in lieu of new field data should high-flow conditions (flow in the overbanks) not occur during the study period. Channel bathymetry data will be collected at the study site and merged with digital elevation data on the flood plains to develop the topographic data set. The goal of the data collection is to acquire two high-flow hydrographic surveys in the study reach for which discharge, cross-section velocity profiles, and water-surface elevations will be collected concurrently. In the absence of two high-flow events, however, two hydrographic surveys will be conducted for varied conditions on a single high-flow hydrograph. Should overbank flow conditions not occur during the study period, consideration will be given to collecting data from median to bankfull flows or using USGS data and models from other complex bridge sites in the southeastern United States to supplement North Carolina data. The hydrographic survey data will be collected as field conditions dictate and will be used to calibrate and validate the models.

A comparison of the output of the 1-D and 2-D models relative to the field data will be the basis for evaluating the differences between the two models at complex riverine and tidal bridges and the implications of these differences. The model comparison also will provide valuable information in the development of guidelines for applying 2-D models at NCDOT bridge sites. The overall investigation is planned to take 2.5 years at which time additional areas of research and application involving multidimensional modeling (with a focus on tidal sites) will be evaluated.