Cracked Girders, November/December 2003 Public Roads

A case study of Case Bridge in Washington, DC, provides some clues about the causes of this kind of structural failure.

Five years ago, a flurry of activity followed the discovery of unexpected cracking in the prestressed girders of the Francis Case Memorial Bridge, an arterial structure spanning the Washington Channel of the Potomac River, in the heart of the Nation's capital. To ensure the safety of the traveling public, the District of Columbia Department of Transportation (DCDOT) with the assistance of the Federal Highway Administration (FHWA) immediately began stabilizing the cracked girders and initiated an indepth investigation to ascertain the cause, prognosis, and whether the structure could be repaired.

Traffic on the Francis Case Memorial Bridge

Francis Case Memorial Bridge crosses the Washington Channel of the Potomac River in Washington, DC.
Photos: Construction Technology Laboratories, Inc.

The Francis Case Memorial Bridge carries eight lanes of I-395 traffic over one channel of the Potomac River in Washington, DC, connecting the downtown with Potomac Park. An extensive rehabilitation program undertaken in 1994 resulted in replacement of the approach spans of the original bridge with precast prestressed concrete girders made continuous at the piers.

During routine inspection of the bridge in 1998, DCDOT and FHWA discovered large, full-depth vertical cracks on the soffit of the beams, near the first interior pier. The cracks in the concrete were of unusual severity and unknown origin.

“Cracking in prestressed elements is undesirable, of course, since it can affect the safety, integrity, and life of the bridge,” notes Joey Hartmann, research structural engineer at Turner-Fairbank Highway Research Center. Hartmann assisted DCDOT with the initial inspection of the problem bridge. Design provisions from the American Association of State Highway and Transportation Officials (AASHTO) prohibit cracking of prestressed concrete structures under service loads.

At the time of the discovery, an FHWA-funded, multiyear applied research study, Jointless and Integral Abutment Bridges, on prestressed girders had just been completed. The data from that study, along with adaptation of findings from prior research on the performance of prestressed concrete bridges, helped assess the cause and severity of the Case Bridge cracking problem. The DC transportation agency retained a consulting firm to conduct a field inspection and analytical evaluation, and the consultant identified the prime cause of distress as the restraint conditions for positive moments at the piers—more on this in a moment.

During the investigation, the consultant identified atypical load cases normally overlooked during design that could be the primary cause for such failures. “These designs are susceptible to the undesirable continuity-induced cracking observed on the Case Bridge,” says Adrian Ciolko, vice president of the consulting firm Construction Technology Laboratories, Inc., of Illinois. He adds, “And counterintuitive to normal design methods, more strength is not always best when designing the continuous zones over supports.”

Background

The multiyear FHWA study, which is being prepared for publication, systematically explored the performance of structures designed according to a widely used but often not clearly understood concept of making concrete and steel simple-span girder bridges continuous for live load. Traditionally, simple-span concrete beams have been made continuous at intermediate supports to serve two key purposes: eliminating joints to reduce maintenance and improve ride quality, and increasing the beam's mid-span capacities for superimposed gravity loads.

The continuity results in inducing negative moments (upward convex deformation) at the intermediate supports due to live load, which typically is addressed by providing reinforcment near the top surface of the cast-in-place diaphragms and slabs at the interior supports.

AASHTO recognizes that under certain types of secondary loading effects, such as creep and shrinkage, positive moments (upward concave deformation) can develop as well. The possibility of positive moments at the interior supports is counterintuitive for most engineers, since most types of loading commonly applied to continuous beam structures typically are gravity-induced and create negative moments at interior supports. Design for positive moments at these locations is generally based on crack control within the diaphragm region of the structure.

Typically, engineers tend to use empirical equations, design charts, standard drawings, or rule of thumb methods for design of the positive moment reinforcement. In the same vein, some may tend to follow the “more is better” philosophy, which suggests that providing more reinforcement than required is considered “conservative.” This practice, however, occasionally can result in unexpected and sometimes detrimental effects, as discovered on the Case Bridge.

Severe Cracking

The Case Bridge consists of five prestressed concrete girder spans over Potomac Park, followed by numerous steel multigirder spans over the Washington Channel of the Potomac River. The five prestressed concrete spans, spanning from the south abutment toward the south edge of the Washington Channel, are designated by letters “A” through “E” and consist of approximately 18 to 20 simple-span standard AASHTO Type III prestressed girders. The simple-span girders were made continuous for live loads via a cast-in-place 216-millimeter (8.5-inch)-thick lightweight concrete deck on stay-in-place forms and a cast-in-place diaphragm. Spans A and B were converted to two-span continuous beams, while Spans C, D, and E were converted to three-span continuous beams.

During a routine inspection, DCDOT observed severe vertical cracks adjacent to the intermediate support at Pier B on eight interior prestressed girders of Spans A and B. The observed cracks were mostly vertical, traversed the complete width of the bottom flange of the girder, and, in some cases, traversed the full girder height.

One of the larger cracks.

Severe cracking and spalling at one of the intermediate diaphragms.

The cracks varied in width from 5 millimeters (0.02 inch) to almost 29 millimeters (1.125 inch) at the concrete's formed surface, with the widest crack located approximately at 1.4 meters (4.5 feet) from the Pier B end of the beam. At the time of the consultant's site visit, the DC Department of Transportation already had shored the cracked girders with steel columns. The outside six girders on either side of the cracked girders showed no visible cracking in the cross-section of the beam. The diaphragm regions of those girders, however, exhibited severe cracking and spalling.

The cracking, especially within the girder cross-section, was of unusual severity and unknown origin, and did not coincide with the commonly recognized distress induced by normal flexural or shear loading. Before attempting to alleviate the problem, however, understanding of why the cracking occurred was essential. The FHWA-funded study on jointless bridges provided a theoretical and experimental basis for uncovering the cause of the problem.

Field Inspection and Measurement of Contributing Issues

The consultant conducted an indepth field inspection and assessment to create a baseline condition profile for the distressed girders and to obtain detailed, specific information about the extent and likely causes of the distress. The field information collected by the consultant included the crack widths, crack lengths, cambers and deflections, temperature gradients, actual creep coefficients, and coefficient of thermal expansion.

Worker inspecting a hydraulic temporary support jack

Closeup of hydraulic temporary support jack.

Specifically, the consultant collected this source data by conducting the following activities:

Evaluation of the Contributing Issues

The analytical evaluation based on the FHWA-sponsored jointless bridge research and the measured values for creep, shrinkage, and differential thermal analyses made it evident that strong potential existed for the observed cracking and distress under specific combinations of differential thermal loads, concrete creep, and shrinkage properties.

One of the less understood and often overlooked issues in design of simple-span prestressed precast beams made continuous for live loads is the effect of secondary moments on the performance of the structure. Converting a simply supported girder—which is without end constraints and is free to deform—to a continuous girder results in the introduction of restraints in the structural system. The result is restraint-induced moments and shears due to loading or environmental effects. These moments and shears commonly are termed “secondary” effects.

The particular load effects of interest for the Case Bridge investigation were the positive secondary moments, that is, those loads causing tensile stresses and potential cracking at the soffit of the girders at intermediate supports. The magnitude of the positive moment is controlled by the amount of positive reinforcement provided at the support diaphragms. By providing a large amount of positive moment reinforcement at the diaphragms, designers inadvertently make the diaphragm area stronger than the adjacent girder sections, thereby forcing the cracking to occur in far more critical but weaker areas of the girder span.

Analytical investigation showed that the large positive moments generated on the Case Bridge were due primarily to the restraint provided by the positive moment reinforcement. “In addition, we surmised that a compromise in bond properties of the prestressing strands or other factors such as presence of lubricants also could have reduced the effective prestressing force at the end of the positive moment reinforcement within the girder cross-section,” says Ciolko, “thereby creating an unusually weak section susceptible to cracking under the applied moments.”

On the girders that did not display cracking along the span, cracking and spalling occurred in the diaphragm regions. “We inferred that inadequate laps in positive moment reinforcement and the cracking and spalling in the diaphragm area resulted in release of the restraint conditions at the supports,” says Dr. Ralph Oesterle, the consulting firm's program manager for the FHWA-funded jointless bridge research study, “thus preventing the cracking from occurring within the spans of those girders.”

Potential Fatigue Concerns

Temperature differentials between the top and bottom of a bridge structure subject the prestressed beams to cycles of varying restraining moments, resulting in the opening and closing of cracks. This action concentrates stresses in the prestressing strands that cross the crack, which can result in two different degradation mechanisms. In the first mechanism, the stress variation can lengthen the debonded surfaces of strands adjacent to the crack surfaces, reducing available prestress in other sections and initiating additional flexural cracks on either side of existing cracks. In that case, the strands will experience lower variation in the stress range. In the other mechanism, the bond may stay intact while the strands may experience higher stress variation and a reduced fatigue life. In either case, the service life of the structure potentially could be shortened.

Conceptualized Rehabilitation Options

Of primary urgency was the immediate shoring of all the girders that were cracked or showed imminent potential for similar distress. At the time of the consultant's inspection, most of the girders already had been shored. The consultant recommended that additional girders susceptible to similar distress should be shored to ensure the immediate safety of the structure. “This approach allowed DCDOT to pursue other more robust permanent rehabilitation options pending availability of sufficient funding,” says FHWA's Hartmann.

Another problem to be resolved before design of a retrofit was the elimination or minimization of the prime cause of the original distress. Although the majority of beams were not cracked, the positive moment detail at the piers had the potential to create substantial restraint in the future, thereby causing cracking similar to what had already occurred.

“In view of this problem, we considered it advisable to recommend modification of the positive moment connection details by eliminating or reducing the number of positive moment reinforcing bars in the diaphragm region,” says Oesterle. The reduction would decrease the positive moment capacity at the diaphragm, thus reducing the magnitude of possible restraint. The crack therefore would form in the positive moment region of the diaphragm, rather than in the girder section.

Diagram of deformation if Left as Simple Spans, Restraint Moment if Made Continuous

Diagram of Positive Restraint Moment and Negative Restraint Moment

These schematics show deformation if spans are left as simple spans, and restraint moments if they are made continuous. The second schematic shos positive and negative secondary moments caused by restraint.

In addition to immediate shoring and modification of the positive moment reinforcement details, the cracked beams will need to be replaced or rehabilitated to ensure public safety in the long term. DCDOT recently initiated a project to explore options to restore the shear and flexural capacity of the affected beams. The scope of the project includes investigating several methods of external post-tensioning using composite fiber wraps, permanent shoring, and eventual replacement of the beams. When the most feasible method is selected, contracting options will be explored to perform the repairs.

In closing, it is important to note that this seemingly simple transformation of simple-span prestressed girders to continuous spans should be attempted with caution, and significant attention must be paid during analysis and design to include loading conditions that can cause counterintuitive behavior such as secondary positive moments at the piers. More importantly, positive moment reinforcement should be designed and detailed such that any cracking, if it occurs, should be limited to the relatively less critical diaphragm region of this type of structural system.

Niket M. Telang, P.E., is a senior engineer with Construction Technology Laboratories, Inc., and was the project manager for the Case Bridge evaluation. He has more than 10 years experience in bridge engineering and has managed several projects in bridge research, inspection, rating, rehabilitation, and design. He is currently the principal investigator for NCHRP Project 10-64, Field Inspection of In-Service FRP Bridge Decks, and managed NCHRP Project 10-43, Movable Bridge Inspection, Evaluation, and Maintenance Manual, in the past. Telang received his B.S. and M.S. degrees in civil engineering from Victoria Jubilee Technical Institute (University of Bombay) and Virginia Polytechnic Institute and State University, respectively.

Armin Mehrabi, Ph.D., P.E., is a senior principal engineer with Construction Technology Laboratories, Inc., and was actively involved in the Case Bridge evaluation. He leads the company's long-span bridge engineering activities, with a focus on the use of innovative techniques for bridge evaluation and inspection. In 1997, Mehrabi was chosen as one of Engineering News-Record magazine's Top 25 Newsmakers for his contribution to the development of nondestructive techniques for evaluation of cable-stayed bridges. He received his master's degree and doctorate in civil engineering from the University of Tehran and the University of Colorado at Boulder, respectively.

November/December 2003

Cracked Girders by Niket M. Telang and Armin B. Mehrabi

Cracked Girders
by Niket M. Telang and Armin B. Mehrabi