HIGHWAY BRIDGES

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

BRIDGE RECONSTRUCTION

Overview

Immediately following the earthquake, Caltrans took measures to reopen obstructed freeway and surface street arteries as soon as possible. Removal of collapsed and/or seriously damaged structures began within one day of the earthquake. Caltrans also quickly decided to rebuild the defunct structures in their former locations with only minor adjustments to their alignments. Design, review, and construction of the replacement structures was expedited. The entire road network was planned to be fully restored by 1 December 1994. New design criteria were evolved for the design of the replacement structures with the assistance of an independent seismic safety review panel convened for the reconstruction. Ten replacement bridges were designed to replace five that had collapsed and five that were damaged. A summary of the replacement bridges is provided in Table 3. Replacement costs are 1994 projections. One bridge, the Truck Connector Overcrossing (Bridge #53-1962F), was repaired, although it was intended initially to be replaced.

Table 3 Northridge Earthquake replacement bridges (Quon 1994)

Design Objectives

The design objectives adopted by Caltrans for the Northridge earthquake replacement structures were intended to prevent collapse under a maximum credible earthquake without explicitly addressing damageability and functionality issues for earthquakes of various magnitudes. In addition, the designs adopted the current Caltrans approach that aims to force inelastic action at predetermined and well-detailed locations in columns while the remainder of the structure remains essentially elastic.

Additional design measures were promulgated for the Northridge replacement structures in response to undesirable behavior observed in this and other recent seismic events. Ductile structures with a high degree of serviceability are expected to result from application of current Caltrans seismic design approaches supplemented by the additional measures described below.

• Minimize the number of expansion hinges to reduce unseating problems.
• Avoid large skews in the plan geometry of abutments, bents and expansion hinges.
• Use prismatic bridge columns throughout the column length.
• Use site-specific ARS spectral curves where available.
• Consider vertical ground accelerations in the design of the bridge superstructure.
• Balance stiffness between bents in individual frames to reduce torsional response of frames and overloading of stiff bents.
• Avoid damage to below-ground structures (except for cast-in-drilled-hole pile shafts) and oversize below-ground structures to accommodate possible future strengthening of the colunns and/or superstructure should future seismic design requirements exceed current requirements.
• Use the latest design criteria as outlined in the draft document of ATC-32 (dated July 1993 subsequently published as “Improved Seismic Design Criteria for California Bridges: Provisional Recommendations” in 1996) for reduction of elastic spectral forces.
• Verify displacement capacities by inelastic push-over analyses based on moment-curvature assessment of inelastic members.

Meeting the above design objectives required a significant change and advance from current Caltrans seismic design practice.

Design Constraints

Design constraints, besides those caused by the short timetable necessary to expedite construction, consisted of:

• maintaining, in most cases, the original road alignment, while
• avoiding existing footings of bents and abutments to the extent possible,
• preparing alternative designs in both steel and concrete for selected bridge structures,
• considering design-build contracts,
• and coordinating simultaneous design efforts and associated seismic safety reviews.

To facilitate the short timetable, cast-in-drilled-hole piles were used where possible and standard Caltrans multicell posttensioned box girders were used in bridge superstructures, except where steel alternates were designed.

Design and Reconstruction

Construction of the replacement bridge projects proceeded at a rapid pace. Six bridges and two structures of the ten replacement bridge projects were completed and opened to traffic less than six months after the earthquake. The remaining replacement bridge projects were scheduled to open before 1 December 1994. The rapid construction of the Northridge replacement bridges is facilitated by financial incentives for the contractor for early route reopening (and disincentives for late reopening) with respect to the contracted period.

CONCLUSIONS

Several important observations can be made regarding the performance of the transportation system. Many of these may be useful in planning future work on that system:

• The effects of the earthquake are not surprising in light of today's knowledge. In this earthquake, all bridge collapses were associated with poor performance of older shear-critical columns and short seat widths, both of which have been identified as critical in previous earthquakes. The progress in knowledge and design practice was evident in the good performance of bridges constructed or retrofitted to current standards. Research and developments in design implementation should continue for improving reliability of modern designs.
• Notwithstanding the progress in knowledge and design practice, this earthquake demonstrated the potential vulnerability of bridges constructed in the period from 1971 (immediately following the San Fernando earthquake) until the mid to late 1970s, a period during which new design force levels and ductile detailing practice were phased into California bridge design practice. Bridges constructed in this era had previously been thought to be of lower priority for evaluation and retrofitting. The Bridge Seismic Retrofit Program may need to place these bridges at a higher priority level.
• The behavior of architectural flares should be examined further. Earlier design practice assumed these flares spalled off, whereas actually they can act as an integral part of the column. Studies should examine (a) actual behavior of column flares detailed according to older and current practice, (b) consequences of flares spalling off or acting monolithically with the column core, (c) effective retrofitting procedures, and (d) effective new design procedures. It is noteworthy that Caltrans had recognized that flares may not perform as intended in the earlier design practices, and had prepared evaluation procedures that would identify this behavior. It is uncertain how uniformly these procedures were implemented.
• Current and older design procedures may not adequately represent the distribution of design forces in long, multiframe bridges. Failure and damage in end frames indicates that the stiffer end frames may attract larger forces than anticipated in design. Some issues involved with this behavior have been identified in recent research and design developments. (a) Pounding at hinges can significantly increase forces and inelastic ductility demands on stiff frames. This increase is not accounted for in current analysis methods. (b) Stiffness distribution is sensitive to assumptions (often difficult to make accurately) on abutment stiffness, column stiffness, foundation stiffness, and location of fixity in shaft foundations. (c) Innovative strategies may mitigate this potential problem, possibly including column articulations, sleeves on cast-in-drilled-hole piles, and seismic isolation bearings.
• As with other construction forms, structural changes can occur during construction and maintenance that invalidate the design assumptions. This seemed to be a factor in channel walls at Bull Creek Canyon Channel UC, amount of overburden on footings at several locations, in-situ conditions of cast-in-drilled-hole pile foundations, and locking of skew collars as at Mission-Gothic UC. It is important to verify that as-built conditions are consistent with design assumptions, and not just the design drawings. Improved documentation of design assumptions may be desirable.
• Phase I hinge restrainer retrofits had mixed performance. There were many examples of restrainer unit failure (cable fracture, fitting failure, and diaphragm punching). More investigation is needed to determine where these restrainers were effective, and to determine improved restrainer designs.
• Older bridges with 6-inch and 8-inch seats should be reevaluated, even if retrofitted with restrainers.
• Phase II retrofits appeared to perform well in all instances. Although this earthquake was of short duration, the high level of accelerations promotes confidence that the retrofits will be successful in a major earthquake.
• Several skewed bridges collapsed. An apparent cause was torsional response that is associated with skewed geometries. Methods are needed for improved design of skewed bridges, possibly including elimination of the skew where feasible, elimination of in-span hinges, and lengthening of seats.
• Damage to abutments and approaches was widespread. Considering the extent of this damage for a moderate earthquake, it seems that the current design strategy to accept abutment damage should be reevaluated.
• Major damage and collapse in multicolumn bridges indicates that the increased reliability associated with redundancy of multicolumn bents is not necessarily sufficient to avoid collapse. Retrofit priorities given to multicolunm bridges and high-traffic-volume bridges may need to be reconsidered. Furthermore, damage and collapse required closure of bridges that usually carry heavy traffic volume.
• Modern traffic management techniques were effective in managing traffic following the earthquake. Furthermore, emergency procedures, including strong incentive and disincentive clauses, were effective in achieving rapid reconstruction of the freeways.

REFERENCES

Astaneh, A., B. Bolt, K. M. McMullin, R. R. Donikian, D. Modjtahedi, and S.-W. Cho. 1994. Seismic performance of steel bridges during the 1994 Northridge earthquake. University of California, Berkeley. Report No. UCB/CE-STEEL-94/01. April.

Brodsly, D. 1981. L. A. Freeway. Berkeley: University of California Press.

Buckle, I.G.ed. 1994. The Northridge, California earthquake of January 17, 1994: Performance of highway bridges. NCEER-94-0008. National Center for Engineering Research, State University of New York at Buffalo, 24 March.

California Department of Transportation. 1976. Report concerning earthquake resistant design advances for state highway and bridge construction. Third Biennial Report. State of California, Business and Transportation Agency, October.

Housner, G. W. 1990. Competing against time. The Governor's Board of Inquiry on the 1989 Loma Prieta Earthquake. G. W. Housner, Chairman, State of California, Office of Planning and Research, 31 May.

Jennings, Paul C. ed. 1971. Engineering features of the San Fernando earthquake of February 9, 1971. California Institute of Technology. Report EERL 71-02, June.

Los Angeles Times. 1994. February 16.

Moehle, J. P. ed. 1994. Preliminary report on the seismological and engineering aspects of the January 17, 1994, Northridge earthquake. UCB/EERC94/01. Earthquake Engineering Research Center, January.

Priestley, M. J. N., F. Seible, and C. M. Uang. 1994. The Northridge earthquake of January 17, 1994: Damage and analysis of selected freeway bridges. University of California, San Diego. Structural Systems Research Project Report No. SSRP 94/06, February.

Priestley, M. J. N., F. Seible, R. Verma, and Y. Xiao. 1993. Seismic shear strength of reinforced concrete columns. University of California, San Diego. Structural Systems Research Project, Report No. SSRP-93/06, July.

Quon, F. 1994. Personal communication, Caltrans District 7, August 17.

Roberts, J. E. 1992. Bridge seismic and other research needs: Caltrans overview. In Proceedings, Third NSF Workshop on Bridge Engineering Research in Progress. La Jolla, California, 16 and 17 November.

Selna, L. G., L. J. Malvar, and R. J. Zelinski. 1989. Bridge retrofit testing: Hinge cable restrainers. ASCE journal of Structural Engineering 1 14 (4): 920-934.

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