Diamonds from the S.C. clay

Nov. 13, 2003

A replacement of the Cooper River bridges is currently under construction and when completed will provide a 1,005-m-long cable-stayed main span over the Cooper River between the city of Charleston and the town of Mount Pleasant in South Carolina. This will be the longest cable-stayed bridge in North America, surpassing the 930-m-long Alex Frasier Bridge in Vancouver, British Columbia.

A replacement of the Cooper River bridges is currently under construction and when completed will provide a 1,005-m-long cable-stayed main span over the Cooper River between the city of Charleston and the town of Mount Pleasant in South Carolina. This will be the longest cable-stayed bridge in North America, surpassing the 930-m-long Alex Frasier Bridge in Vancouver, British Columbia.

The bridge evolved from studies begun in 1988 to address the need to replace the deficient existing Silas T. Pearman and Grace Memorial bridges between Charleston and Mount Pleasant and to address a more recent need to improve shipping clearances in the upper reaches to the Charleston harbor, the second busiest port on the east coast of the U.S. In order to address traffic and shipping demands, the new bridge will provide a 30-m horizontal and 57.5-m vertical clearance over the main channel and a 76-m horizontal and 20-m vertical clearance over a secondary channel, Town Creek.

The two 175-m-high diamond-shaped towers carry a 43.3-m-wide composite deck with eight traffic lanes and a 3.6-m pedestrian walkway and bikeway on the south edge. The total suspended span length is 1,005 m, with a main span of 471 m and back spans of 198 m and 69 m on each side. It will utilize a composite concrete deck with steel edge girders. The sidewalk/bikeway will be cantilevered outside of the south edge girder.

The high-level approaches also will utilize composite steel construction with steel girders spaced 3.66 m on centers. The high approaches are jointless and 1,326 m long on the Charleston side and 637 m long on the Mount Pleasant side. Beyond the high-level approach spans there are low-level approach spans and interchange structures. These utilize composite precast concrete girders for the straight portions and composite steel girders for the curved ramps.

The environmental conditions at Charleston are probably among the most challenging in the U.S., involving strong winds and earthquakes. During the Great 1886 Charleston Earthquake, church bells rang from Cuba to Boston. In addition, Charleston is on the South Carolina coast, an area prone to hurricanes and the associated strong winds.

The entire site is characterized by a layer of stiff clay known as Cooper Marl at a depth of 15-18 m below Elevation 0. Above the marl, the river has soft alluvial deposits, while the land portions of the project have relatively soft surficial soils, so the bearing stratum throughout the site is the Cooper Marl.

On shaky ground

Charleston is well-known for the 1886 magnitude-7.3 earthquake. Recent studies that have led to the development of a project-specific seismic-design criteria suggest that during the past 2,000-5,000 years up to six large earthquakes of similar magnitude have occurred, and they appear to have originated from the same source.

The seismic-design criteria developed for this project employed the performance-based two-level design earthquake approach. For the more severe 2,500-year-return-period earthquake, the main span and designated portions of the approach structure could not suffer significant damage and would need to be able to be returned to service shortly after such an event. The remaining portions of the structure could suffer considerable damage but could not collapse.

For an event with a return period of 500 years, the structure was to remain in the elastic range. The solution to these demanding criteria in a very competitive design-build environment--with a client that had only a limited budget--proved to be the major design challenge.

In general the solution adopted, particularly for the main span and high-level approach spans, was to provide sufficient ductility of the structures while minimizing the structure weight to provide enough flexibility in the structure so that seismic demands were minimized.

The use of 3-m-diam. drilled shafts in the foundations both minimized the number of shafts and provided the adequate support capacity. For the main span piers only 11 of the high-capacity drilled shafts are required, while on the high-level approaches typically only two drilled shafts are required per pier.

The height of the high-level approach span piers proved to be too flexible along the axis of the bridge so the solution adopted was to make the approach spans continuous over a significant length so the shorter piers of the lower portions of the approaches act to brace the taller piers. This has an advantage of minimizing expansion joints but has required considerable analysis to demonstrate its feasibility.

The project's seismic-design criteria required a time-history analysis using records from three different events as part of the final design of these portions of the bridge.

A site-specific design spectrum was prepared using three scaled earthquakes: Tabas, Imperial Valley and Joshua Tree. Seismic design of the main span and the adjacent high-level approaches utilized response spectrum analysis (for preliminary design), push-over analysis and full inelastic nonlinear time-history analysis.

A final check of the seismic performance of the main span was made using inelastic time-history analysis of the main span and the west and east high-level approach structures combined in a single model. This model, representing almost 3,000 m of bridge, had 55,350 degrees of freedom, 14,316 elastic elements, 4,460 inelastic elements and 563 sets of ground motions.

Nonlinear inelastic time-history analysis confirmed that the structure is adequate for the design earthquakes.

Columns built to stand

The main span anchor and side piers were designed for non-seismic loads in accordance with the AASHTO Strength Design Method and checked for serviceability requirements. The design forces for non-seismic loads were taken from a global nonlinear elastic analytical model of the main span (using the GTStrudl program).

Once the longitudinal reinforcement of the columns was determined based on the non-seismic loads, inelastic static analysis (push-over analysis) of the stand-alone piers was performed. The displacement demands for the push-over analysis were taken as 1.5 times the elastic displacement demands from the response spectrum analysis of the global main span model.

For each pier, separate push-over analyses were performed in the transverse and longitudinal directions. In each direction separate push-over analyses were performed using two sets of material strengths: maximum feasible material strengths and expected material strengths.

The design forces for the shear reinforcement in the pier columns are taken as the peak shear demands from the push-over analyses--pushed to failure, not just to 1.5 times the elastic displacement demands (as required by the project seismic design criteria).

The push-over analyses using the maximum feasible material strengths controlled with respect to shear demands on the columns and force and moment demands on cap beams and intermediate struts. The push-over analysis using the expected material strengths controlled for the displacement capacities. In all cases the displacement capacities are greater than 1.5 times the elastic displacement demands.

Push-over analyses also were performed for the main pylons. However, the complex geometry and mass distribution of the pylons was such that the push-over analyses for the pylons yielded results which were questionable at best.

We concluded that push-over analysis is a very useful and expedient design tool for structures whose response is predominantly in a single mode, but it is not very useful for structures whose response cannot be adequately represented by a single mode. Therefore, in order to ensure that the pylons will yield in flexure before failing in shear, shear design of the pylon legs was checked directly against the ADINA time-history analysis.

The seismic analysis uses motions in two horizontal directions combined with vertical motions. A site-response analysis with SHAKE (a computer program) is used to establish ground motions.

An evaluation of available references found that if adequately detailed, hollow piers would perform well.

The seismic-design criteria indicate hinging should be prevented below grade in the foundation elements. Because the bridge is founded on large-diameter drilled shafts--the shafts vary from 3 m diam. on the main span to 1.8-2.4 m diam. on the low-level spans--the design of the drilled shafts has in general been controlled by the plastic hinge capacity of the columns above the drilled shafts. Since most of the shafts are in 15-18 m of soft soil, the shafts have had their maximum moment at the level of the stiff marl just below the soft upper soil layer. This has resulted in having more reinforcing steel at the mid-portion of the drilled shaft than at the top, a somewhat unusual arrangement but one dictated by the site conditions.

Rock islands

Because Charleston is a major port and has plans to increase the present 15-m-deep channel to 18 m, an important element of the design was ship collision. The bridge currently sees 3,900 transits per year. The criteria for the new bridge required projecting traffic to 2050, about midway through the life of the structure.

As part of the ship protection scheme, rock islands founded on the marl will protect the main piers against ship collision.

On the approach piers that are in deep water, the piers have been strengthened to provide adequate protection so that the return period for the collapse of the bridge due to ship collision is 1 in 10,000 years.

Avoiding a big blow

The wind engineering study involved four factors: (a) wind climate analysis including site measurements; (b) selection of design criteria; (c) stability, comfort and wind load appraisal for the completed bridge; and (d) wind loads and stability of the critical construction stages. The following presents these studies undertaken in the laboratories of Rowan Williams Davies & Irwin Inc., Guelph, Ontario.

For the wind climate analysis, historical wind data were obtained from the Charleston International Airport (from 1945 to 1999). The surrounding terrain was assessed through site visits, topographic maps and photographs. Hurricane simulations and meteorological data from other sources also were used.

For short return periods such as 10 years, the dominating winds were found likely to be from the south, southeast or east. These winds were attributed mainly to normal wind events. For longer return periods such as 100, 1,000 or more years, hurricanes were the main contributor with northeast being the dominant wind direction. The mean speed profiles and turbulence properties were estimated using empirical methods.

For the design of the completed bridge 48.3 m/sec (1 m/sec = 2.25 mph) and for construction 33.5 m/sec mean hourly wind speeds were recommended. These speeds correspond to 100- and 10-year return periods, respectively, at deck height.

The objective of the wind tunnel testing was to ensure an aerodynamically stable, comfortable and economical bridge. The bridge deck was tested as a sectional model at scale 1:80, considering various barriers, railings and fairings, until satisfactory configurations were determined.

The most favorable modification included a half-diamond fairing on the north edge and sloped soffit covering the underside of the sidewalk on the south edge. To improve the aerodynamic stability and minimize the wind loads, this modified section was retained over the middle part of the main span, whereas for the rest of the deck, the original section was kept.

Employing the final deck configuration, the stability of the completed bridge was verified on a 1:250 aeroelastic model. The modified deck configuration was installed in the middle of the main span over a total length of 236 m, whereas the remaining portion of the main span did not include any aerodynamic modifications.

A race to the top

The foundations for the new Cooper River bridge are now substantially complete, according to Charles Dwyer, the project manager for the South Carolina Department of Transportation (SCDOT). Bobby Clair is the director of engineering for the SCDOT, which is the owner of the bridge. The construction contractor is Palmetto Bridge Construction.

On the upward construction effort, the builders have reached the anchorage zone of the towers, the top 150 ft where the stay cables eventually will be attached. On the downward effort they "bottomed out" in September. Dwyer told Roads & Bridges, "We reached substantial completion of our foundations." There are about nine shafts yet to be drilled under one of the ramps, but other than that ramp, they have completed all the drilled shafts. "That's a pretty good milestone for us."

Overall, the project is about 60% complete. The contractor recently started pouring the concrete deck on top of the girders of the approach spans. One piece of equipment they are using is a 147-ft screed, the longest ever built by Bid-Well.

The deck over the steel-girder, cable-stayed spans will be precast and erected in balanced cantilever fashion. Some of the sections have already been cast in Savannah, Ga., and are undergoing a six-month cure before being shipped to the bridge site.

They plan to start work on the pier table, the section adjacent to the towers where there are no cables anchored, by the end of the year. Then they will start erecting the precast sections.

About The Author: Abrahams, Bryson and Wahl are engineers at Parsons Brinckerhoff Quade & Douglas Inc., New York. Stoyanoff is an engineer at Rowan Williams Davies & Irwin Inc., Guelph, Ontario, Canada.