HIGHWAY BRIDGES

Abridged from Earthquake Spectra,
Vol 11, Supplement C, Chapter 6,
April, 1995
Used by permission of EERI

DETAILED DESCRIPTIONS OF FREEWAY BRIDGE DAMAGE

Interstate 5/State Road 14 Interchange

The I-5/SR14 Interchange is located approximately 12 km north-northeast of the epicenter. The interchange consists of a number of bridges of various lengths. Four major multispan bridges connect the SR14 to northbound and southbound I-5. A truck route extends to the southeast of I-5 predominantly on grade, creating a need for a number of smaller bridges. Figure 2 shows an oblique view of the interchange and identifies the bridges.

The interchange was severely damaged in the 1971 San Fernando earthquake (Jennings 1971). The interchange was under construction during the 1971 earthquake, when a portion of the south I-5/east SR14 connector collapsed. This connector was the highest bridge; pier lengths were as high as 165 feet. Construction after the earthquake was completed using the original plans, except new designs were prepared for the few columns that had not been constructed prior to the earthquake and for those that required reconstruction. The spacing of transverse reinforcement in the new designs was reduced. Restrainers were added at expansion hinges during the final stages of construction.

The interchange was severely damaged by the Northridge earthquake; portions of the south I-5/west SR14 connector and the north I-5/west SR14 connector collapsed (Fig. 3). These structural failures are described in more detail in the following sections. Pounding damage was observed at expansion hinges and abutments of standing portions of these structures and other structures at the interchange, indicative of large relative structural displacements.

Ground Motion. The site is located in a relatively narrow valley bordered by steep hills. No accelerograms were obtained from the site or immediate vicinity. Mean attenuation relations suggest that peak horizontal accelerations were about 0.5g at the site. Topographic amplification and out-of-phase motion at the periphery of the valley may have affected the motions incident on the structures.

Route 14/I5 Separation and Overhead (Bridge #53-1960F)

Bridge Description. This ten-span curved box-girder bridge connected westbound traffic on SR14 to southbound I-5. The bridge was constructed as five frames separated by expansion hinges. Prestressed concrete box girders were used in the frames at each end of the bridge and in the central frame. The second and fourth frames, separating the prestressed portions, were constructed with conventional reinforced concrete. Plan and elevation drawings are shown in Figure 4. The five-cell box girder was 55 feet wide by 7 feet deep; deck and soffit slabs were 7 inches and 6 inches thick, respectively. A cross section of the deck is shown in Figure 5.

The design prestress force balanced approximately 60% of dead load. From construction documents and reconnaissance observations, it appeared that the prestressing ducts had not been grouted.

Single-column bents were used throughout the bridge; column cross sections were rectangular with bull-nosed ends. Columns were reinforced with #18 Grade 60 longitudinal bars and #4 Grade 40 transverse hoops and cross ties at 12-inch centers, as shown in Figure 6. The number of longitudinal bars varied from column to column. At Bents 2 and 3 in Frame 1, which collapsed, 40 #18 column bars had been provided. These columns were supported by 12-foot-diameter drilled shafts terminating approximately 40 to 60 feet below grade. Above grade, column heights varied over the length of the bridge from about 27 feet at Bent 2 to 107 feet at Bent 7. Seat-type abutments with external wingwall shear keys were used at each end of the bridge. The abutments were founded on spread footings.

The site is in a relatively narrow valley bordered by steep hills. Bases of the pile shafts are supported in highly weathered sandstones and silty sandstones of the Towsley formation. Surface conditions depend on grade changes made during the construction of the I-5/SR14 Interchange. More than 40 feet of cut at Abutment 1 and Bent 2 revealed very firm and laterally stiff soils. At other locations along the bridge, where cuts were smaller or fill was placed, the lateral stiffness of the foundation may have been lower.

Observed Damage. Frame 1, extending from Abutment 1 to the expansion hinge in Span 3, collapsed, as shown in Figure 7. The column at Bent 2 was reduced to rubble. The box girder fractured over the column at Bent 3; adjacent spans fell to either side of the column (Figure 8). The east wingwall at Abutment 1 was severely damaged, presumably by impact of the box girder. The box girder lost seat support and moved about 5 feet north of the abutment face (Figure 9). Span 3 of Frame 1 lost seat support at the expansion hinge; restrainer and equalizing bolts at the expansion hinge were fractured. The fallen portion of Span 3 was displaced to the north of its original position in plan, laying partially beneath Frame 2.

Figure 7	Figure 8	Figure 9

Failure Assessment. The fracture of the box girder at Bent 3 gave the appearance that the Bent 3 column had punched through the bridge superstructure. This led to early speculation that the failure was attributable primarily to vertical accelerations. Subsequent evaluations (Priestley et al. 1994; Moehle 1994) have discounted this theory. Preliminary calculations (Priestley et al. 1994) indicate extremely high accelerations (sufficient to have thrown the spread footing abutments into the air) would have been required to shatter Pier 2 in compression or to fail the box girder at Pier 3.

It is more plausible that the collapse was primarily the consequence of horizontal ground motion and subsequent response of the bridge. As the bridge responded to the ground motions, the column at Bent 2, which was stiffer than the other bents, attracted a high proportion of the inertial loads. This column failed in shear, and consequently lost its axial load capacity. The box girder did not have sufficient strength to span from Abutment 1 to Bent 3, and failed in flexure at Bents 2 and 3. The collapse of the box girder pulled Span 1 north of Abutment 1. Fracture of the prestressing tendons at Bent 3 released sufficient energy to fracture the top 12 inches of the bent cap, as may be seen in Figure 10. The integrity of the connection between the box girder and the bent cap was lost, and Spans 2 and 3 fell to either side of Bent 3. Rotations associated with the failure at Bent 3 were sufficient to fracture the bar restrainers at the expansion hinge in Span 3. Span 3 of Frame 1 lost seat support at the expansion hinge. As it fell, it rotated about its area of contact with the ground at Bent 3, clipping the base of the hinge of Frame 2 and landing partially beneath it.

Mission-Gothic Undercrossing (Bridge #53-2205)

Bridge Description. SR118 is carried over Mission Boulevard and Gothic Avenue on two parallel bridge structures designed in 1973. The surface streets intersect at right angles just to the northeast of the bridge. For each structure, the abutments were oriented parallel to the adjacent streets; the abutments were skewed at approximately 45 degrees to the longitudinal axis of the bridge and were orthogonal to each other (Figure 11). The right structure carried eastbound SR118 traffic toward the southeast. The right structure was supported on three 2-column bents. The left structure, which was not as long, was supported on two 2-column bents. Bents 2 and 3 of the right structure (identified as 2R and 3R) and Bent 2 of the left structure (2L) were aligned parallel to Mission Boulevard. Bents 3L and 4R were offset approximately 25 degrees from the alignment of Gothic Avenue.

The superstructures consisted of 10- or 11-cell box girders. The 7.5-footdeep box girders were prestressed longitudinally along the girder lines and transversely along the cap beams. Typical superstructure dimensions are shown in Figure 11.

The column height from the top of the footing to the bridge soffit varied from 22.2 feet to 24.7 feet, as shown in Figures 12a and 12b. Column cross sections were six-foot octagons flaring over the upper 12 feet in the bent direction to a rectangular section measuring 14 feet by 6 feet. Each of the ten columns was reinforced with 45 #11 Grade 40 longitudinal bars that continued through the flare in a circular arrangement. Spiral reinforcement consisted of smooth Grade 40 #5 spiral at a 3.5-inch pitch; the spiral was spliced via butt welds with backing bars. Flare reinforcement consisted of 22 #11 bars embedded six feet into the cap beam and terminated approximately at the base of the flare. Rectangular stirrups were provided over the upper eight feet of the flared section, consisting of #5 bars at 12-inch centers. The lower four feet of the flare was not enclosed by transverse stirrup reinforcement.

Plan dimensions of the rectangular footings ranged between 20 feet and 24 feet in both directions; footing depth ranged between 5 feet and 6.5 feet. Each footing had top and bottom reinforcing mats and was supported by between 25 and 36 concrete piles rated to 70 tons. These concrete piles were 16 inches in diameter. To accommodate movements associated with the posttensioned superstructure, each column was supported on greased neoprene and sheet metal layers at the foundation level. Movement was allowed for more than 120 days after prestressing operations, and then a reinforced concrete collar was cast around the column against a 2-inch-thick layer of expansion joint filler. The collars were doweled into the footing. Collar dimensions and reinforcement details are shown in Figures 12a and 12b.

A five-foot-wide seat was provided at each abutment, and the ends of the superstructure were supported on twelve 18-inch by 18-inch by 3-inch elastomeric bearing pads. Three 28-inch square shear keys were provided at each bridge end. The shear keys were embedded 15 inches into the abutment seat; six inches of movement capacity was provided parallel to the bridge alignment.

Observed Damage. A large portion of the right structure collapsed and the superstructure of the left structure settled one to two feet at Bent 3L (Figure 13). Movement was predominantly toward the southwest along the alignment of Bents 3L and 4R, almost perpendicular to the bridge alignment.

Figure 13	Figure 14	Figure 15

Two damage patterns were observed in the columns. These patterns were associated with the orientation of the bent. Bent 3L (Figure 14) featured a failure of a plastic hinge at the bottom of the column flare. The #5 smooth spiral reinforcement fractured at several locations, and the main longitudinal reinforcement buckled where the spiral fractured. Second, the flared regions of the columns at Bents 2L, 2R, and 3R, which were oriented parallel to Mission Boulevard, were heavily spalled. There were signs of flexural hinging at the top of the column and shear distress over the entire flare length (Figure 15). Flare damage appeared to have begun where the longitudinal flare reinforcement was no longer enclosed by stirrups, and to have propagated up as the concrete cover spalled over the flared region.

The right column of Bent 2R moved laterally at ground level. A 1- to 2-foot gap was observed on one side of the column; an earthen bulge had developed at the other side. Subsequent excavation by Caltrans indicated that the shear collar cracked and likely had failed, allowing the column to slide on the footing. The column was offset approximately 1.5 feet from the center of the footing (California Department of Transportation 1994).

At Abutment 1R, there were displacements greater than 1.5 feet parallel to the abutment and approximately 0.5 feet parallel to the bridge alignment. Most of the displacement appeared to have been a consequence of the collapse of other portions of the bridge, but the magnitude of displacement suggests that the abutment shear keys also must have failed. Also, significant cracking was observed at shear key locations in the abutment seat at Abutment 5L.

Failure Assessment. Superstructure displacements of both bridges were guided toward the southwest by the cant abutments. Modes of failure observed in standing portions of the bridge are thought to be indicative of possible failure modes in collapsed portions of the right structure.

Column flares influenced failure behavior. In columns displaced predominantly along the bent axis (the "strong" direction of the column), a well-defined plastic hinge formed at the base of the flare. Columns loaded in this direction were vulnerable to shear failure as a consequence of the short shear-span length resulting from the presence of the flare. Fracture of the spiral reinforcement, buckling of the longitudinal bars, and distortion of the plastic hinge region in the direction of the bent suggest that confinement levels were inadequate to secure ductile flexural behavior.

Columns loaded predominantly in the "weak" direction showed concrete spalling over the entire flare length. Because moment demand and capacity increase concomitantly toward the top of the flare, the long region of spalling suggests the development of a long plastic hinge zone in these columns. Diagonal cracks in the flare are indicative of shear degradation in this hinge zone.

Because the least deformability is expected for flared columns loaded in the "strong" direction, the collapse of portions of the right structure can be attributed to failure of Bent 4R loaded along the bent axis. Damage to other bents of the right structure occurred by loading predominantly in the "weak" direction of the flared columns. The failure in Bents 3L and 4R may be responsible for the level of damage observed in Bents 2L, 2R, and 3R.

Updated December 17, 1997.
Send comments or questions to EERC Library