Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 165
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council 6 Highway Bridges James E. Roberts INTRODUCTION The California State Department of Transportation (CALTRANS) owns and maintains over 12,000 bridges with spans over 20 feet. There are an equal number in the city and county systems. CALTRANS maintains the condition data for all of these and some 6,000 other highway structures such as culverts (with spans under 20 feet), pumping plants, tunnels, tubes, highway patrol inspection facilities, maintenance stations, toll plazas, and other transportation-related structures. Structural details and the current condition data are maintained in the Department Bridge Maintenance files as part of the National Bridge Inventory System required by Congress and administered by the Federal Highway Administration. These data are updated and submitted annually to the Federal Highway Administration and are the basis upon which some of the federal gas tax funds are allocated and returned to the states. The maintenance, rehabilitation, and replacement needs for bridges are prorated against the total national needs. Only this year (1993) has seismic retrofitting been accepted as an eligible item for use of federal funds, because it was assumed by most other states to be only a California problem. After much lobbying by CALTRANS, the new Federal Intermodal Surface Transportation Efficiency Act of 1992 provides for seismic retrofit to be eligible for federal bridge funds. Immediately after the February 9, 1971, San Fernando earthquake, CALTRANS began a comprehensive upgrading of their Bridge Seismic Design Specifications and Construction Details. CALTRANS's bridge design specifications
OCR for page 166
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council were modified to correct the identified deficiencies for application on new bridge designs. After this work was completed, the Applied Technology Council completed project ATC-6, which became the basis for a similar seismic-design specification for bridges, which was adopted by the American Association of State Highway and Transportation Officials as the national standard in 1983. Existing structures, however, proved to be a substantially more challenging problem. Research was undertaken in the United States and overseas (in New Zealand and Japan) to improve analytical techniques and to provide basic data on the strengths and deformation characteristics of lateral-load resisting systems for bridges. CALTRANS identified the vulnerable elements of existing bridges and began a statewide seismic-retrofit program for bridges to systematically reinforce the older, non-ductile bridges. The initial phase of the CALTRANS Bridge Seismic Retrofit Program involved installation of hinge and joint restrainers to prevent deck joints from separating. Separation was the major cause of bridge collapse during the San Fernando earthquake and was judged by CALTRANS engineers and other investigators to be the highest risk to the traveling public. Included in this phase was the installation of devices to fasten the superstructure elements to the substructure in order to prevent those superstructure elements from falling off their supports. This phase was essentially completed in 1989 after approximately 1,260 bridges on the state highway system had been retrofitted at a cost of over $55 million. Funding for this program competed with other highway safety programs, which were arguably more critical in terms of statistical support. Consequently, the bridge seismic retrofit program was allocated only $4 million annually. While the hinge and joint restrainers performed well, shear failure of columns on the I-605/I-5 separation bridge in Los Angeles during the moderate Whittier earthquake of October 1, 1987, reemphasized the inadequacies of pre-1971 column designs. Even though there was no collapse, the extensive damage resulted in plans for basic research into practical methods of retrofitting bridge columns on the existing pre-1971, non-ductile bridges. That research program had already been initiated in early 1987 at the University of California (UC), San Diego, and the Whittier earthquake merely speeded its approval and execution. Funding levels for seismic-retrofit-program implementation were increased fourfold after the Whittier earthquake to an annual level of $16 million. Even at that level, it would require some 100 years to complete the retrofitting program that is currently identified. The Loma Prieta earthquake of October 17, 1989, again proved the reliability of hinge and joint restrainers, but the tragic loss of life at the Cypress Street Viaduct on I-880 in Oakland emphasized the necessity to immediately accelerate the column-retrofit phase of the seismic-retrofit program for bridges with a higher funding level for both research and implementation. Other structures in the earthquake-affected counties performed well, suffering the expected column
OCR for page 167
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council damage without collapse. With the exception of a single outrigger column-cap joint confinement detail, those bridges using the post-1972 design specifications and confinement details performed well. Damage to long, multiple-level bridges showed the need to consider more carefully longitudinal resisting systems, because earthquake forces cannot be carried into abutments and approach embankments as they can on shorter bridges. After the Loma Prieta earthquake caused 44 fatalities on the state highway system, capital funding for seismic retrofitting was increased to $300 million per year. At the same time, seismic-research funding for bridges was increased from $0.5 million annually to $5.0 million annually with an initial $8.0 million allocation from the special State Emergency Earthquake Recovery legislation of November 1989, Senate Bill 36X (SB 36X). Using the special research funding provided in SB 36X, the department engaged additional research teams and facilities to assist in this massive program. Much research has been conducted by both U.S. and foreign researchers into the causes of damage in the Loma Prieta earthquake, and much of that research is contained in the references cited in this paper. Most of the research papers can be obtained from the National Information Service for Earthquake Engineering, Earthquake Engineering Research Center at UC Berkeley. The Earthquake Engineering Research Center, located at the Richmond, California, Field Station, has been designated as the national repository for information on the Loma Prieta earthquake. There are over 175 documents on file at the repository relating to bridge aspects of the Loma Prieta earthquake. Additional research papers and project reports can be obtained from the CALTRANS Division of Structures, Sacramento, California; the Department of Applied Mechanics and Engineering Science, UC San Diego; and the National Center for Earthquake Engineering Research, State University of New York at Buffalo. The National Center for Earthquake Engineering Research has a data base search service known as QUAKLINE. PERFORMANCE OF PRIOR RESEARCH RESULTS Much had been learned about bridge performance in previous earthquakes (e.g., the 1971 San Fernando and 1987 Whittier earthquakes), and only budgetary constraints prevented CALTRANS from executing seismic retrofit of older bridges at a more rapid pace. It is important, however, to observe and discuss the performance of the new seismic-design criteria that had been utilized on bridges designed after 1972 and those seismic retrofit devices that had been installed prior to the Loma Prieta event. Hinge-Joint Restraining Devices As previously stated, the initial phase of CALTRANS' Bridge Seismic Retrofit Program involved installation of hinge and joint restrainers to prevent deck joints from separating. This was identified as the major cause of bridge collapse during the San Fernando earthquake (LeBeau et al., 1971) and in 1972 was
OCR for page 168
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council judged by CALTRANS engineers to be the highest risk to the traveling public. Included in this phase was the installation of devices to fasten the superstructure elements to the substructure in order to resist vertical accelerations and also to prevent superstructure elements from falling off their supports. This phase was essentially completed in 1989 after approximately 1,260 bridges on the state highway system had been retrofitted at a cost of over $55 million. Research and testing of the restrainers were conducted at UC Los Angeles, by Selna and Malvar (1987). These joint restrainer systems have performed well in subsequent earthquakes, including the 1987 Whittier event (Priestley et al., 1991c), the 1989 Loma Prieta event (Mellon et al., 1993), the 1992 Cape Mendocino event (Yashinsky, 1992a), and the three most recent southern California events of 1992. In the eight counties that were declared disaster areas after the Loma Prieta earthquake, there are approximately 350 bridges that had been retrofitted with hinge joint restrainers. There was no observed failure of any of these restrainers. CALTRANS staff engineers agree that there would have been collapse of bridge spans due to the spans falling off their supports without the installation of restrainers. Maragakis and Saiidi (1991) of the University of Nevada at Reno and Yashinsky (1990) of the CALTRANS Office of Earthquake Engineering have published papers evaluating the performance of these restrainer details. The University of Nevada at Reno was awarded a CALTRANS research project (Project R-12) to test the performance of hinge and joint-cable restrainers for bridges under dynamic loading. Properly Confined Column Reinforcement Most columns designed since 1971 contain a slight increase in the main-column vertical reinforcing steel and a major increase in confinement and shear reinforcing steel over the pre-1971 designs. All new columns, regardless of geometric shape, are reinforced with one or a series of spiral-wound interlocking circular cages. The typical transverse reinforcement detail now consists of #6 (.75-inch-diameter) hoops or continuous spiral at approximately 3-inch pitch over the full column height. This provides approximately eight times the confinement and shear reinforcing steel in columns than what was used in the pre-1972 non-ductile designs. All main-column reinforcing is continuous into the footings and superstructure. Splices are mostly welded or mechanical, both in the main and transverse reinforcing. Transverse-reinforcing steel is designed to produce a ductile column by confining the plastic hinge areas at the top and bottom of columns. The use of grade 60, A 706 reinforcing steel in bridges has recently been specified on a few projects on a trial basis. In the eight counties declared disaster areas after the Loma Prieta earthquake, there are approximately 800 bridges designed after 1972 using the newly revised seismic-design criteria and confinement details. With the exception of the one outrigger beam-column joint damage on the I-980 southbound connector
OCR for page 169
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council in Oakland, there was no documented damage to any of these 800 post-1972-designed bridges (Mellon et al., 1993). Acceleration Response Spectra For Alluvium And Dense Foundation Materials CALTRANS developed a series of acceleration response spectra for alluvium and average soils after the 1971 San Fernando earthquake, and these spectra were accurate for prediction of the dynamic response of those types of foundation materials. Professor Harry Bolton Seed of UC Berkeley was instrumental in the development of these design spectra. Those bridges situated on average foundation materials and designed using these spectra performed well in the Loma Prieta event. Base-Isolated Girder Systems Although only one bridge in the affected counties was base isolated, it did perform well during the Loma Prieta event (Mellon et al., 1993). The bridge was the Sierra Point Overhead, which was designed prior to 1972 for lateral-force requirements of only 0.06 g. It was subjected to lateral forces of approximately 0.18 g during the Loma Prieta earthquake and showed no signs of distress. It should be noted, however, that the CALTRANS design procedure is to force seismic loads into the abutments so the back wall must fail prior to the base-isolation bearings being engaged. PROBLEMS WITH EXISTING CRITERIA, DETAILS, AND PRACTICE A discussion of the problems encountered in highway bridge performance during the Loma Prieta earthquake will explain the need for research in the area of structural response in moderate and major earthquakes. Older Bridges Designed For Pre-1972 Seismic Forces And Design Criteria The major causes of bridge damage in the Loma Prieta earthquake were the criteria and details for which they were originally designed. There were over 4,000 bridges on the combined state, county, and city systems in the eight counties that were declared disaster areas after the earthquake. Only 100 of those bridges were damaged in the earthquake, and only 25 sustained what can be termed major damage, as reported in the Post Earthquake Investigation Team (PEQIT) Report (Mellon et al., 1993). Only one of the 800 bridges in the counties that had been designed after 1972, using the newer seismic forces and details, suffered damage as described in the PEQIT Report (Mellon et al., 1993).
OCR for page 170
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council While the Loma Prieta earthquake was, admittedly, a moderate earthquake, the bridge performance was generally what had been expected by bridge designers. Most of the research that has been commissioned since the earthquake is aimed at developing better assurance that bridges will withstand a major earthquake without collapse or major damage and that the transportation system can remain essentially functional after a major seismic event. Seismic Performance Criteria Required The Governor's Board of Inquiry hearings brought out the fact that there was no formal documented policy on the required seismic performance of bridges in the CALTRANS Design Specifications and Criteria (Housner, 1990). These specifications are utilized by many other public agencies, and, therefore, it is critical that a formal performance criteria be adopted. Dynamic Response Of Deep, Soft Foundations The effect of the dynamic response of deep, soft soils in the structure foundations also proved to be a contributing factor to the collapse of the Cypress Viaduct and must be analyzed and included in future design procedures, especially for long, tall structures with relatively high periods of vibration. The effect of incoherence in the foundation response is also an important factor in the design of very long structures such as the San Francisco-Oakland Bay Bridge and the mile-long freeway viaducts. Mitchell (1992), Bolt (1991, 1992), Der Kiureghian (1991), Der Kiureghian and Neyenhofer (1992), Zafir et al. (1990), and Tamura and Shah (1991) have published research papers on this subject. Column-Footing Interaction Investigation of damage at the Cypress Street Viaduct in Oakland subsequent to the Loma Prieta event revealed a deficiency in many pre-1972-designed bridge footings. Some of these footings suffered joint shear failures that caused structure settlement. These footings were typically designed for vertical loads and only a 0.06 g lateral force. Subsequent investigation and research by Seible and Priestley of UC San Diego revealed a potential for failures due to lack of reinforcing steel in the top to resist lateral moments. Their conclusions, based on analysis and tests, did show a need for a top mat of reinforcing steel (Seible et al., 1992a). Inadequate Column-Confinement Reinforcement Other than the Cypress Viaduct failure, column damage was limited to a few critical bents on the Embarcadero Freeway Viaduct, the Terminal Separation
OCR for page 171
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Structure, the Central Freeway Viaduct, and the Southern Freeway Viaduct (Route 280) (Mellon et al., 1993). Generally, those damaged bents were located in areas over deep, soft soil and bay mud. The damage on the Central Viaduct was located in a few bends on the northern end between Oak and Turk streets. This was the only damage to portions of a structure that was not constructed over deep, soft soils. These structures were closed almost immediately with the exception of that portion of the Central Viaduct south of Oak Street, where there was no sign of damage. Temporary splice beams were installed on those columns of the Central Viaduct where column hinge joints had been located in the original design. This splice was intended to keep the joint from separating in a future seismic event until a more permanent retrofit detail with new columns could be installed. A report of the analysis and recommendations was prepared for CALTRANS by Seible and Priestley (1991). The most spectacular damage and that which was closest to collapse occurred in the vicinity of Innis Street on the Southern Freeway Viaduct, Interstate 280. The shorter of two columns supporting long outrigger bents failed in joint shear near the lower deck level. This occurred at only four bent locations on the structure, however. While the damage was minimal, there was obvious concern for the integrity of these pre-1972 design, non-ductile, reinforced-concrete structures. They had all been designed in the late 1950s to early 1960s for lateral forces of 0.06 g, using details of the period that we now know were weak and provided insufficient confinement, especially at beam-column joints. All the damaged areas were shored up with heavy-timber falsework to reinforce them during aftershocks and possible future seismic events until permanent repairs could be made. Since the duration of the Loma Prieta earthquake was relatively short and the magnitude moderate, it was prudent to close the structures to public traffic until they could be retrofitted to current seismic standards. This damage has been reported in Cooper and Van de Pol (1991), Elsesser and Whittaker (1991), Fenves (1992), Miranda and Bertero (1991), Moehle et al. (1991), Priestley and Seible (1990), Seible and Priestley (1991), and Thewalt and Stojadinovic (1992). Inadequate Beam-Column Joint Reinforcement Research and analysis conducted subsequent to the Loma Prieta earthquake have shown conclusively that the lightly reinforced column-pedestal detail unique to the Cypress structure was the main cause of the total collapse. Immediately following the earthquake, the Structural Engineering Department at UC Berkeley requested that CALTRANS's Division of Structures salvage a section of the damaged Cypress Street Viaduct that had not collapsed for the purpose of conducting a series of seismic performance experiments. UC Berkeley has published reports of the experiments that were to determine the fundamental period of the complex structure under low-intensity seismic loading. Additional tests
OCR for page 172
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council were performed to measure the actual lateral resistance of the framing system up to yield. Experiments on several proposed column-retrofit details were also conducted on this structure. Bollo et al. (1990), Miranda and Bertero (1991), Mahin and Moehle (1990; Mahin, 1991), Kay (1991), and Jones and Schroeder (1991) have published results of this analysis and research. Inadequate Torsional Reinforcement The China Basin Viaduct suffered bent outrigger damage at two locations in the vicinity of the Sixth Street northbound off ramp. That ramp was closed to traffic, but the mainline structure was kept open since it was a single-level structure on multiple-column bents, and no damage was observed at any other location. Emergency contracts were awarded to shore up the damaged bent and to effect a complete replacement of both the damaged bents and also the supporting columns at one bent location. This work was completed while mainline traffic was allowed to continue operating on the structure. Subsequent analysis of the outrigger performance indicated shear failure and a potential for combined torsion and shear failure, as reported by Moehle (1992; Moehle et al., 1991). Pedestrian-Bridge Performance During the vulnerability screening and seismic analysis of bridges subsequent to the Loma Prieta event, the CALTRANS Office of Earthquake Engineering staff has noted that the class of single-column-supported pedestrian bridges have universally required seismic-retrofit strengthening. They are generally lightweight and very narrow, offering little lateral stiffness, thus rendering them especially vulnerable to seismically induced lateral forces. Steel-Bridge Performance The documented failures of structural steel bridges during the earthquake were few, and they can be attributed to the date that those bridges were designed. The anchor-bolt failures on the San Mateo-Hayward Bridge, the San Francisco-Oakland Bay Bridge, and the viaducts on the east end of the Bay Bridge were major failures, but they were repaired in short order, especially the anchor bolts on the San Mateo-Hayward Bridge and on the East Bay Distribution Structure. Subsequent investigation revealed a large number of structural steel columns on the San Francisco Skyway portion of Interstate 80 and US 101 that must be strengthened to prevent collapse because of the low seismic lateral forces for which they were designed. There was no evidence of structural steel-girder failure on any of the bridges in the affected counties.
OCR for page 173
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Importance Factor Applied To Critical Bridges The Governor's Board of Inquiry, in its report of June 1, 1990, recommended that ''The Department of Transportation should adopt a seismic safety policy for transportation structures that assures that transportation structures are seismically safe and that important transportation structures maintain their function after earthquakes'' (Housner, 1990). During hearings conducted by the Board of Inquiry, the board discussed this issue with CALTRANS engineers. There had never been a distinction between bridges relative to their importance to the state or local community. As a direct result of the one-month loss of the San Francisco-Oakland Bay Bridge during the Loma Prieta earthquake, it has been recommended that major transportation structures be designed to remain essentially elastic for higher seismic-force levels and longer shaking periods to reduce the damage to a non-structural type. To accomplish this goal, a new "importance factor" was introduced into the design and retrofit performance criteria. This represents a major change in the seismic-design criteria for bridges and also represents the introduction of a subjective factor that will be based on judgment more than engineering principles. RESEARCH IN SEISMIC RESPONSE OF BRIDGES As a result of the problems discussed above and in response to further direction by the Governor's Board of Inquiry, which recommended that "The Department of Transportation should fund a continuing program of basic and problem-focused research on earthquake engineering issues pertinent to CALTRANS responsibilities" (Housner, 1990), and the Governor's Executive Order D-86-90, June 2, 1990, the Department of Transportation immediately accelerated its Bridge Seismic Research Program with the funding of 23 major research projects at a total cost of over $8 million. This research was "problem focused" on those areas that proved to be vulnerable during the recent earthquakes. Most of the research involved half-size model testing of bridge components and joint details. Seismic-retrofit details to strengthen existing bridges were developed and proven with this research program and their good performance in three recent earthquakes in California in April and June 1992, prove the validity of the ductile design and retrofit approach adopted by CALTRANS. After the Loma Prieta earthquake caused 44 fatalities on the state highway system, capital funding for seismic retrofitting was increased from $4 million to $300 million per year. At the same time, seismic research funding for bridges was increased from $0.5 million annually to $5 million annually with an initial $8 million allocation from the special State Emergency Earthquake Recovery legislation of November 1989, Senate Bill 36X. Research has been conducted, and is currently underway, at the UC San
OCR for page 174
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Diego, to test and confirm the validity of several proposed design solutions for seismic retrofitting of existing single-column-bent substructure elements on bridges. Since the Loma Prieta earthquake, additional research has been and is currently being conducted at the University of California at Berkeley, San Diego, Irvine, and Davis to develop and test retrofit techniques for multiple-column bents and double-level structures, including abutment and footing details. Development of Vulnerability-Analysis Algorithm In order to set priorities for more than 24,000 bridges in the state for order of seismic-retrofit upgrading, CALTRANS engineering staff developed a risk-analysis procedure and adjusted it over the next three years as more information became available. Identification of bridges likely to sustain damage during an earthquake was an essential first step in the Single-Column phase of the Bridge Seismic Retrofit Program, which had begun just prior to the Loma Prieta earthquake. What can be classified as a level-one risk analysis was employed as the framework of the process that led to a consensus list of risk prioritized bridges. This risk analysis procedure was later utilized to also prioritize the multiple-column-supported bridges, but the single-column-supported bridges were deemed more vulnerable, based on experiences during the 1971 San Fernando earthquake. A conventional risk analysis produces a probability of failure or survival. This probability is derived from a relationship between the load and resistance sides of a design equation. Not only is an approximate value for the absolute risk determined, but relative risks can be obtained by comparing the determined risks of a number of structures. Such analyses generally require vast collections of data to define statistical distributions for all, or at least the most important, elements of some form of analysis, design, or decision equations. The acquisition of this information can be costly if obtainable at all. Basically, what is done is to execute an analysis, evaluate both sides of the relevant design equation, and define and evaluate a failure or survival function. All of the calculations are carried out, taking into account the statistical distribution of every equation component designated as a variable throughout the entire procedure. To avoid such a large, time-consuming investment in resources and to obtain results that could be applied quickly as part of the Single Column phase of the retrofit program, an alternative was recognized. What can be called a level-one risk-analysis procedure was used. The difference between a conventional and level-one risk analysis is that in a level-one analysis judgments take the place of massive data-supported statistical distributions. The level-one risk-analysis procedure used can be summarized by the following steps: Identify major faults with high event probabilities (priority-one faults). This step was carried out by consulting the California Division of Mines and
OCR for page 175
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Geology and recent U.S. Geological Survey studies. A team of seismologists and engineers identified faults believed to be the sources of future significant seismic events. Selection criteria included location, geologic age, time of last displacement (late quarternary and younger), and length of fault (10-km minimum). Each fault recognized in step 1 was evaluated for style, length, dip, and area of faulting in order to estimate potential earthquake magnitude. Known faults were placed in one of three categories: minor (ignored for the purposes of this project), priority two (mapped and evaluated but unused for this project), or priority one (mapped, evaluated, and recognized as immediately threatening). Develop attenuation relationships at faults identified in step 1. An aver-age-attenuation model was developed by Mualchin and Jones (1992) of the California Division of Mines and Geology to be used throughout the state. It is the average of several published models. Define the minimum ground acceleration capable of causing severe damage to bridge structures. The critical (i.e., damage-causing) level of ground acceleration was determined by performing nonlinear analyses on a typical, highly susceptible structure (single-column connector ramp) under varying maximum ground-acceleration loads. The lowest maximum ground acceleration that demanded the columns provide a ductility ratio of 1.3 was defined as the critical level of ground acceleration. The level of ground acceleration determined in this study was 0.5 g. Identify all the bridges within high risk zones defined by the attenuation model of step 2 and the critical acceleration boundary of step 3. The shortest distance from every bridge in California to every priority-one fault was calculated. Each distance was compared with the distance from each respective level of magnitude fault to a 0.5 g decrement acceleration boundary. If the distance from the fault to the bridge was less than the distance from the fault to the 0.5 g boundary, the bridge was determined to lie in the high-risk zone and was added to the screening list for prioritization. The prioritization procedure is described below. The CALTRANS Division of Structures has developed a computerized data base that has the coordinates of all 24,000 state, county, and city bridges stored. CALTRANS can produce a map of the entire state or any portion of the state showing the bridges, the major faults, and an overlay of the combinations. These maps can be viewed on the computer screen or printed for use by designers in screening to identify high-risk bridges. The procedure is quite simple using the computer data base to locate all highway bridges on the state system, locate all earthquake faults, then determine those structures that are in a high-risk zone. Prioritize the threatened bridges by summing weighted-bridge structural and transportation characteristic scores. This step constitutes the process used to prioritize the bridges within the high-risk zones to establish the order of bridges to be investigated for retrofitting. It is in this step that a risk value is assigned to each bridge.
OCR for page 208
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Priestley, M.J.N., and F. Seible. 1993. Assessment and Testing of Column Lap Splices for the Santa Monica Viaduct Retrofit. In Proceedings: ASCE Structures Congress XI, Irvine, California, April. Ramey, M.R., et al. 1991. Experimental Testing of Epoxy Injected and Steel Shell Retrofitted Sections from the Collapsed Struve Slough Bridge. In Proceedings of the First Annual Seismic Research Workshop, California Department of Transportation, Sacramento, California. Rashid, Y.R., R.A. Dameron, and I.R. Kurkchubasche. 1992. Predictive Analysis of Outrigger Knee-Joint Hysteresis Tests: A Torsion/Flexure Tests at University of California, San Diego. Report to University of California, Berkeley, University of California, San Diego, and CALTRANS , Sacramento, California. Roberts, J.E. 1989. Bridge Seismic Retrofit Program For California Highway System. In Proceedings: U.S.-Japan Workshop on Lifeline Earthquake Engineering, Public Works Research Institute, Tsukuba Science City, Japan, May. Roberts, J.E. 1989. Theory of California Seismic Bridge Design And Analysis For The Beginner. California Department of Transportation, Division of Structures, Training Course for Entry Level Engineers, July. Roberts, J.E. 1990. Recent Advances in Seismic Design and Retrofit of Highway Bridges. In Proceedings: Earthquake Engineering Research Institute Annual Meeting Palm Springs, California, May. Roberts, J.E. 1990. Recent Advances in Seismic Design and Retrofit of Highway Bridges. Structural Engineers Association of California, Proceedings: 59th Annual Meeting, Incline Village, Nevada. September. Roberts, J.E. 1991. Recent Advances in Seismic Design and Retrofit of Highway Bridges. Proceedings. Seismic Retrofit Workshop, University of California at San Diego, July. Roberts, J.E. 1991. Seismic Performance of Steel Bridges. Structural Steel Fabricators Annual Meeting, Saint Louis, Missouri, September, 1991. Published in Modern Steel Construction magazine, July 1992. Roberts. J.E. 1991. Recent Advances in Seismic Design and Retrofit of Highway Bridges. In Proceedings of the Third U.S. Conference, Report: Technical Council on Lifeline Earthquake Engineering, Monograph 4. American Society of Civil Engineers, New York, August. Roberts. J.E. 1991. Seismic Retrofitting of San Francisco Viaducts. In Proceedings: 60th Annual Meeting, Structural Engineers Association of California, Palm Springs, California, October. Roberts. J.E 1992. Seismic Design of Bridge Foundations. In Proceedings, Transportation Research Board Annual Meeting, Washington, D.C., January. Roberts. J.E. 1992. Large Scale Model Testing for Seismic Design and Retrofit, Initiating, Funding and Managing. Presented at the Earthquake Engineering Research Institute Annual Meeting, San Francisco, California, February. Roberts, J.E. 1992. Research Based Bridge Seismic Design and Retrofit Program, Criteria, Standards and Status. In Proceedings: Fifth U.S.-Japan Workshop on Earthquake Disaster Prevention for Lifeline Systems, Tsukuba Science City, Japan, October 26. Roberts, J.E. 1992. Effect of Foundation Sod Response on Bridge Seismic Performance. Proceedings: Fifth U.S.-Japan Workshop on Earthquake Disaster Prevention for Lifeline Systems, Tsukuba Science City, Japan, October 26. Roberts, J.E. 1992. Bridge Seismic And Other Research Needs-CALTRANS Overview. In Proceedings: Third NSF Workshop on Bridge Engineering Research in Progress, University of California at San Diego, November 15. Roblee, C.J. 1992. Synthesis of CALTRANS. Foundation Seismic Research Program. Internal Report: Division of New Technology, Materials & Research, California Department of Transportation, Sacramento, California, November. Saadeghvazri, M.A. 1990. Response of the Struve Slough Bridge under the Loma Prieta Earthquake. In Proceedings: Second Workshop on Bridge Engineering Research in Progress, University of Nevada at Reno.
OCR for page 209
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Schnabel, P.B., J. Lysmer, and H.B. Seed. 1972. SHAKE, A Computer Program for Earthquake Response Analysis of Horizontally Layered Sites. Report: Earthquake Engineering Research Center, University of California, Berkeley. Seim, C., and S. Rodriguez. 1993. Seismic Performance and Retrofit of the Golden Gate Bridge. In Proceedings: ASCE Structures Congress XI, Irvine, California, April. Sikorsky, C., N. Stubbs, and M. Richardson. Nondestructive Damage Assessment of a Bridge Using Modal Testing and Structural Reliability. Stuart, R.J. 1991. Seismic Modeling Parametric Studies. In Proceedings of the First Annual Seismic Research Workshop, California Department of Transportation, Sacramento, California. Thorkildsen, E. 1992. Overview of CALTRANS' Bridge Seismic Research Program. Structure Notes, No. 27, California Department of Transportation, Sacramento, California, July. Werner, S.D., and C.E. Taylor. 1990. Seismic Risk Considerations for Transportation Systems. Recent Lifeline Seismic Risk Studies, Report: Technical Council on Lifeline Earthquake Engineering, Monograph 1, American Society of Civil Engineers, New York, November. Werner, S.D., S.A. Mahin, and N.C. Tsai. 1993. Compilation and Evaluation of Current Bridge Damping Data Base. Report to CALTRANS, February. Werner, S.D., L. Katafygiotis, and J. Beck. 1993. Seismic Analysis of Meloland Road Overcrossing Using Calibrated Structural and Foundation Models. In Proceedings: ASCE Structures Congress XI, Irvine, California, April. Whittaker, A.S., E. Elsesser. 1991. Seismic Design Criteria for Transportation Structures. Lifeline Earthquake Engineering: Proceedings of the Third U.S. Conference, Report: Technical Council on Lifeline Earthquake Engineering, Monograph 4, American Society of Civil Engineers, New York, August. Zelinski, R. 1990. California Highway Bridge Retrofit Strategy and Details. In Proceedings: 59th Annual Convention, Structural Engineers Association of California, Sacramento, California. September. Zelinski, R. 1990. California Highway Bridge Retrofit Strategy and Details In Proceedings: Second Workshop on Bridge Engineering Research in Progress, University of Nevada at Reno. Zelinski, R. 1991. San Francisco Double Deck Viaduct Retrofits. Lifeline Earthquake Engineering: Proceedings of the Third U.S. Conference, Report: Technical Council on Lifeline Earthquake Engineering, Monograph 4, American Society of Civil Engineers, New York, August. Zelinski, R., and A.K. Dubovik. 1991. Seismic Retrofit of Highway Bridge Structures. Lifeline Earthquake Engineering: Proceedings of the Third U.S. Conference, Report: Technical Council on Lifeline Earthquake Engineering, Monograph 4, American Society of Civil Engineers, New York, August. COMMISSIONED RESEARCH PROJECTS R-1. Retrofitting of Bridge Columns, Stage I, using third to half scale model tests. UC, San Diego. Contract $400,000. Completed June 30, 1990. This is the first contract with Dr. Priestley which was initiated in early 1987. The emphasis was on single columns using steel jackets for confinement. The physical testing is complete and all reports are completed. R-2. Guidelines for Effective use of Nonlinear Structural Analysis for Bridge Structures. UC, Berkeley. Contract $ 47,000. Completion July, 1993. Principal investigator Dr. Graham Powell. This project is for providing guidelines and training to the Department in the use of the new non-linear analysis programs recently installed. Work is underway. R-3. Shear Strength Capacity Vs. Rotation of Column Pins at Base of Elevated Roadway Structure, using third to half scale model tests. UC, Irvine. Contract $80,000. Completion date April 30, 1993. Principal Investigator Dr. Robin Shepard. This project is necessary because of the many columns which are pinned at the base on multi-column bents. As we move into the multi-column phase of seismic retrofit, we need this information. Report will be submitted spring 1993.
OCR for page 210
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council R-4. Retrofitting of Bridge Columns, Stage II, using third to half scale model tests. UC, San Diego. Contract $331,000. Completion June 30, 1991. This is the second contract with Dr. Priestley and is a continuation of the previous work using different retrofit techniques and different column types. Work is completed and reports are submitted. R-5. Experimental Testing of Epoxy Injected Steel Shell Retrofitted Sections From Collapsed Struve Slough Bridge, using full-scale tests. UC, Davis. Contract $53,000. Completion date June 30, 1993. Principal investigator Dr. Melvin Ramey. We salvaged several broken piles from the Struve Slough Bridge at Watsonville for this test. Since we have a large number of these type bridges in the state, both state and locally owned, it is necessary to test techniques for improving their seismic performance. Work completed. Reports will be submitted spring 1993. R-6. Evaluation of the Dumbarton Bridge in the Loma Prieta Earthquake. UC, Berkeley. Contract $50,000. Completion date June 30, 1991. Principal investigator Dr. Gregory Fenves. The Dumbarton Bridge was instrumented with strong motion instruments and the records are available. It is the only long crossing of the San Francisco Bay which was designed to modern bridge earthquake codes and criteria and this performance evaluation will be useful to the profession. Work completed. Report submitted. R-7. Seismic Response of Deep Soil Sites in the San Francisco Bay Area. UC, Berkeley. Contract $315,000. Completion date December 31, 1992. Principal investigators Dr. John Lysmer, Dr. Raymond Seed. As a major part of the comprehensive earthquake vulnerability evaluations of important transportation structures we have a need to first determine the foundation response. This research is a direct result of problems with structures constructed on deep, soft soils during the Loma Prieta earthquake. The results of this project will be a new set of Acceleration Response Spectra (ARS curves) for deep, soft soils. Work completed. Report due spring 1993. R-8. Retrofitting of Bridge Columns, Stage III, using third to half scale model tests. UC, San Diego. Contract $768,000. Completion July 12, 1993. This is the third contract with Dr. Priestley and is a continuation of the previous work using different retrofit techniques and different column types. This project, however, concentrated on the multiple column bridge bent configuration where the columns typically have moment connections at both top and bottom, but in many cases are pinned at the base. Work completed. Report due spring 1993. R-9. Seismic Retrofit of Bridge Column Footings, using third to half scale model tests. UC, San Diego. Contract $374,000. Completion July 12, 1993. This is the fourth contract with Dr. Priestley and is the final contract m the series to evaluate and test the best techniques to retrofit the footings of older bridges where the original design moments introduced into the footings were much smaller than now anticipated after columns are retrofitted to carry more moment. Work underway. R-10. Evaluation and Retrofitting of Multi-Level and Multi-Column Structures, using third to half-scale model tests. UC, Berkeley. Contract $1,900,000. Completion December 31, 1993. Principal investigator, Dr. Stephen Mahin. This research is essential for the future evaluation of older and newer multi-level structures. Interim results will be used to confirm the techniques and details being used in the current retrofit program. Work underway. R-11. Develop Bridge Seismic Design Criteria with Higher Degree of Safety and Reliability than Provided with Current Design Procedures. Applied Technology Council. Contract $800,000. Completion date is December 15, 1993. This project is to evaluate the current CALTRANS/AASHTO bridge design criteria/code and recommend improvements and changes. R-12. Evaluation of the Performance of Bridge Cable Restrainers During the Loma Prieta Earthquake. University of Nevada, Reno. Contract $91,000. Completion date June, 1993. Principal investigator, Dr. M. Saiidi. Work complete. Final report due spring 1993.
OCR for page 211
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council R-13. Reduced Scale Tests of Pier Walls Under Cyclic Loading for Seismic Retrofit. UC, Irvine. Contract $461,000. Completion date July 1992. Principal investigator Dr. Robin Shepard. Because of the many bridge piers of this design we need test results to determine the best retrofit technique and to provide better design criteria for new pier wall designs. Work completed. Report due spring 1993. R-14. Development of High Strength Fiber Composite Column Wrap. Fyfe Associates, Inc. Contract $73,000. Completion date December, 1991. This is an alternative method to provide confinement on existing bridge columns. It has been used in Japan for industrial smoke stacks using carbon fiber, which is ten times the cost of other fiber composites. Work complete. Report submitted. R-15. Inspection and Data Collection on the Substructure of the Bay Bridge. UC, Berkeley. Contract $49,900. Completion date not yet negotiated. The principal investigator, Dr. Astaneh needed additional funding to complete work begun under an NSF grant. Work complete. Report included with major contract on SFOBB. R-16. Flexural Integrity of Column/Cap Connections Using Number 18 bars. UC, San Diego. Contract $500,000. Completion date December 1992. The principal investigator, Dr. Priestley will build and test full size components to test the bond development length required for these large bars. Most large bridge columns in California utilize these bars and there is some question regarding the adequacy of the AASHTO code requirements. Work complete. Draft report submitted. R-17. Seismic Evaluation of the San Francisco-Oakland Bay Bridge. GENESYS. Proposed contract amount $160,000. This proposal is for a rapid evaluation of the west bay spans using current technology. Has been evaluated by the Seismic Research Advisory Panel. Contract negotiations underway. R-18. Evaluation of Earthquake-Induced Cyclic and Permanent Ground Displacements for Soil Sites. Earth Mechanics. Proposed contract amount $50,000. Still being evaluated. R-19. Development and Implementation of Improved Seismic Design and Retrofit Procedures for Bridge Abutments. University of Southern California. Proposed contract amount $100,000. Completion date not yet determined. Still being negotiated. R-20. Experimental Measurements of Bridge Abutment Stiffness and Strength Characteristics. UC, Davis. Contract amount $350,000. Completion date June, 1993. Principal investigator, Dr. Karl Romstad. This project will incorporate the centrifuge to test models of various combinations of bridge abutment-soil interaction. Work still underway. Report due late spring 1993. R-21. A Simplified, Verified Procedure to Analyze Soil-Pile Structure Interaction During Earthquake Loading Conditions Using an Effective Stress Method. UC, Davis. Contract amount $25,000. Completion date December 1991. Principal investigator Dr. I.M. Idriss. Draft Report due spring 1993. R-22. Response of Pile-Supported Bridge Elements due to Liquefaction. National Cooperative Highway Research Program (NCHRP). Contract cost will be borne by NCHRP if this project is approved. It has national application. Being evaluated by research committee. R-23. Seismic Condition Assessment of the Bay Bridge. UC, Berkeley. Contract amount $800,000. Contract completion date June 30, 1993. Principal investigator, Dr. Abduhollah Astaneh. This project is the major effort to model and conduct an extensive dynamic analysis of the bay bridge with time history to evaluate its response to a larger earthquake such as an 8.0 on the San Andreas or a 7.3 on the Hayward fault. We have decided to begin with the San Francisco-Oakland Bay Bridge for obvious reasons. This project will include a comprehensive analysis of the foundation material response and upon completion we plan to retrofit the bridge to withstand the forces and movements recommended. The project is the beginning of the assessment of all major bay and river crossings in California. Work completed. Draft report submitted. Final report due June 1993.
OCR for page 212
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council R-24.Seismic Hazard Risk Analysis for the San Francisco Bay Region. This proposed project will evaluate the seismic hazard risks for the region and provide appropriate ground acceleration input for geotechnical evaluations of specific structure sites. Geomatrix Consultants completed work. Final report delivered February 1993. R-25. Seismic hazard risk analysis for Southern California. This proposed project will evaluate the seismic hazard risks for the region and provide appropriate ground acceleration input for geotechnical evaluations of specific structure sites. Consultant, Woodward Clyde Associates. Contract completion date, June 1993. Work completed. Draft report due June 1993. R-26. Seismic Condition Assessments of Remaining Toll Bridges. These projects will be advertised by the RFP process and appropriate principal investigators and consultants will be selected for the 9 remaining major toll crossings on the state highway system. They are the Dumbarton Bridge, the San Mateo-Hayward Bridge, The Richmond-San Rafael Bridge, the Carquinez Bridge (1927 and 1955 structures), the Benicia-Martinez Bridge, the Antioch Bridge, the Terminal Island Suspension Bridge (Vincent Thomas Bridge). the Gerald Desmond Bridge on Terminal Island (soon to be taken into the state highway system) and the San Diego-Coronado Bridge. R-27. Implement Advanced Soil Structure Interaction Techniques for the Analysis of Bridge Structures. Consultant, Coast Analytics, Incorporated. Contract completion April 1993. Contract amount $88,500. Work underway. Report due May 1993. R-28. Response of Soft Soil Sites Using the Centrifuge. UC Davis. Various Site Conditions will be Tested at Varying Levels of Shaking to Augment Recordings From Recent Earthquakes and to Calibrate Analytic Procedures. Contract Amount $125,000. Contract Completion Date October, 1993. Principal Investigator Dr. I.M. Idriss. R-29. Construction of Shaker for the Large Centrifuge. U.C. Davis. Partial Support for Construction of This Shaker for Future Research Involving Soil-Pile and Soil-Structure Interaction, Liquefaction, Site Improvement and other projects. Contract Amount $82,500. Contract Completion Date November, 1993. Principal Investigator Dr. I.M. Idriss.
OCR for page 213
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council DISCUSSANTS' COMMENTS: BRIDGES James D. Cooper, Federal Highway Administration We owe a lot to CALTRANS for the way they reacted to the earthquake—namely, in their openness in allowing investigators to come in, examine, and quantify the damage and in their dissemination of information to researchers. This is the chief reason we learn from earthquakes. My interest stems from what the lessons from the Loma Prieta earthquake mean nationally. For example, in California, there are 24,000 bridges; to date, 1,300-1,400 of those have been retrofitted. Nationally, there is an inventory of 577,000 bridges with as many as 75 percent of those being bridges at risk, either because of location or no design for seismic resistance, and few have been retro-fitted. We have learned lessons from previous earthquakes—the San Fernando event being one of the most significant events for the bridge community. Lessons from previous events point to the need to define and accommodate forces, accommodate displacements, evaluate ground-motion amplification and attenuation, and identify liquefaction potential from site-specific studies or from macro analysis. We have also learned that retrofit enhances performance. Loma Prieta was a moderate earthquake, with a short duration of strong ground motion. It produced two to three cycles of inelastic response. We need to consider carefully whether the earthquake can be used as a base for design of structures in other areas of the country, particularly those east of the Rockies, or whether eastern events will produce larger cycles of inelastic response. The latter will have significant impact on the design/retrofit philosophy. Nevertheless, there were technical lessons learned from the Loma Prieta event: Simple retrofit helps. Following the San Fernando earthquake, CALTRANS embarked on a program to identify vulnerable bridges and details and implement a simple, relatively inexpensive retrofit technique to provide displacement control across expansion joints. The relatively good performance of those bridges retrofit with hinge restrainers is testimony that large numbers of structures can be economically retrofit to enhance seismic resistance. Detailing is critical. Bridges designed and constructed following the San Fernando earthquake performed relatively well. Column and column-connection details were revised to accommodate earthquake-generated shears, moments, and pull-out forces. In addition, the development of improved analytical technology was key to improved detail design. Vulnerability assessment is required. Identification of hazard exposure, coupled with site and structural analyses are required to determine structural vulnerability. Only when the system as a whole is compared can a rational decision be made as to which structures—and to what extent—to retrofit.
OCR for page 214
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Mitigation Strategy Design new structures to current criteria. The increased cost associated with providing seismic resistance in newly designed and constructed bridges can vary significantly. However, typical cost increases are about 2-3 percent, a very affordable figure. If all new (or replacement) bridges were designed by the latest standards, it would take upwards of 100 years to reduce the seismic vulnerability of the highway system. In the long term, this strategy is affordable and will prevail. Retrofit the most important/critical bridges. Since most retrofit is very costly, retrofit only those structures that are defined as important with simple, relatively inexpensive technology. Retrofit only those bridges that are defined as critical with the more complex (thereby costly) technology as it evolves. Multidisciplinary, approach required. The Loma Prieta earthquake taught us that scientific and engineering knowledge is advancing; many public policy, legal, and financial issues remain to be resolved; public consciousness and awareness about the catastrophic consequences of a great earthquake are being raised, and emergency planning and response procedures are advancing. It is clear that a sound earthquake-hazard-mitigation strategy will require the coordinated and cooperative involvement of professionals of varied disciplines. Initiatives Required Awareness. Continue promoting technical and public awareness programs. Evaluation. Complete seismic evaluation of the bridge inventory. Design update. Develop philosophically consistent design criteria. Retrofit criteria. Develop and adopt a rational retrofit criteria. In conclusion, earthquake-hazard-mitigation is a long-term endeavor. As we implement new technologies, we are making a significant impact on improving the seismic performance of our highway system. Gregory Orsolini, Parsons Deleuw, Inc. The I-280 Southern Freeway viaduct in San Francisco is a project I am intimately familiar with, having been working on it since just days after the Loma Prieta earthquake. This is a structure on a very high acceleration site and the poorest soils in the Bay Area. We were asked to adhere to the basic philosophy of CALTRANS—to prevent collapse—but also to have a serviceability mechanism addressed. The project was to replace the columns, the joint areas, and the foundations. I'll be concentrating on the reconstruction of these joints and outriggers (some in excess of 50 ft in length).
OCR for page 215
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Beam-Column Joint Shear Design Current design practice for beam-column joint shear for bridges results in heavily reinforced connections. Regarding beam-column joints in bridges, there are some unique considerations—the tendency to use some sizeable bar sizes (#14 and #18 are not uncommon) in retrofit and new construction, and to use large joint shear stirrups. This leads to complications for placement because of the consideration of the bend diameters. We are still understanding/learning about density of reinforcement in the joints. In building these joints, we are fighting uncertainty in the construction industry about how these things are put together. As the contractors learn what we're trying to do, the learning curve should accelerate quickly. To get adequate implementation, we are interacting with contractors and inspectors on how this is actually done. Bent Cap Outrigger Design The original design used a column that was pinned at the top and the bottom, which reduced the lateral bending in these outrigger caps to zero. Our peer review panel asked us to increase redundancy. We did that by adding a fixed joint at the bottom, which adds lateral bending problems to the design of outriggers. That is where we got the flared configuration. For vertical loads, we have post-tensioning in a parabolic shape but also post-tensioning following a bit of a flare. We added bolsters within the box girder to take the reaction of those large flares. We approached the combination of vertical and lateral bending by trying to keep the strains and the pre-stressing strains to below the proportional limit (.008 strain). To consider the vertical accelerations, we multiplied the vertical dead-load by 50 percent—either increasing or decreasing the vertical load by that amount. Use of Simplified Nonlinear Analysis Methods for Seismic Analysis There are many options for nonlinear analysis available. I would like to encourage the use of a more simplified type of analysis, as it has a lot of benefits and is a good tool to use. A number of useful nonlinear analysis methods are currently being used for assessing the ductile behavior of existing structures—displacement ductility, equal energy, and equal displacement concepts can be applied to bridge-retrofit work. It does not necessarily have to be a complex procedure to use nonlinear analysis. Thank you very much.
OCR for page 216
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council Nicholas F. Forell, Forell/Elsesser Engineers After the Loma Prieta earthquake, it was determined that all of the six San Francisco double-deck freeways were in need of retrofit. Most of the freeways (the exceptions were the Alemani Viaduct and the Terminal Separation structure) were damaged and closed to traffic. These structures were built in the late 1950s and have similar construction to the Cypress Viaduct, which failed catastrophically. Immediately after the earthquake on November 1, 1989, CALTRANS retained six major engineering firms with adequate staffing to immediately design and implement repairs and retrofits. The Governor's Board of Inquiry, in its hearings, concluded that because there was no precedent for this type of retrofit, CALTRANS should initiate a peer review process to assure compliance of the designs with the performance criteria established. The peer review panel selected by CALTRANS consisted of six practicing engineers experienced in seismic design and four technical advisors. The technical advisors (two professors from UC Berkeley and two from UC San Diego) were extremely important because of their in-depth knowledge of analysis and concrete design as well as their research experience. The technical advisors supplemented the technical knowledge of the panel members and the consultants. An early and detailed establishment of the scope of work of the peer review panel is important. In this case, the panel reviewed the seismic design criteria and the geologic-hazard report, the design and performance criteria, applicable and available research results, analysis and modeling assumptions, design details, construction documents, and constructability. The panel did not perform a check on the drawings' calculations; therefore peer review cannot be considered a substitute for independent plan checks. It is most important that the peer review panel be convened at the very beginning of the project. Although the consultants were on board immediately after the earthquake, the peer review panel, for many reasons, could not be convened until March 1990. This resulted in wasted effort and money. Some of the work under construction had to be halted or abandoned and much of the design redone. It is therefore important to have the peer review panel functioning as the retrofit concepts are formed. On unique and complex projects, the peer review process can be laborious. On the double-deck retrofit projects, 56 hearings were held. Yet, the peer review process proved itself invaluable and provided CALTRANS with the assurance that the retrofit design will meet the established performance criteria.
OCR for page 217
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council John Clark, Anderson Bjornstad Kane Jacobs I would like to address what worked. We've seen a great deal of spectacular things that failed, but it is important to address what did work—anything designed to the current code (i.e., the post-1983 guidelines that Mr. Cooper alluded to—the ATC-6 document). The ATC-6 document was a revelation that we are dealing with a displacement phenomena. Under the old—even the post-1971 San Fernando codes—we simply doubled our forces and went about things the same way. The code we have now is a good code. I don't mean to say that it solves all our problems. The current code has brought a strong need to consider the foundation effects—the liquefaction potential. Unfortunately, in building bridges, we don't have the luxury to choose the best sites. Consideration should be given to soil-structure interaction effect—not only as it may tend to amplify things but also as it may tend to reduce the response through damping and other effects. The strength of the current code is its concentration on providing ductile details and in providing adequate seat length. Mr. Cooper made an important point about the efficacy of simple retrofits—the joint restrainers. The empirical design tools we have for that are very crude, but they work. Seismic isolation is another very promising tool. There was not anything in the bridge field during the Loma Prieta earthquake that gave us a good test, but it is a good principle. There are things we need to learn about it yet—particularly the increased vulnerability at the joints and the means to address that. We also need to focus our research on whether we can relax some of the confinement provisions we have in the code. That's a constructability issue more than anything else. There are tools out there that we need to get into our code—such as a reduction factor based on our percentage of axial load. We have not yet incorporated that into the code because people think they are being conservative. I would beg to differ with that. A few brief comments on what I might call the retrofit ''philosophy'': There is a critical need to think clearly about what we are doing in codifying the retrofit process in general, because it is a very cost-benefit-sensitive issue. The need for retrofitting far outweighs the available resources. Do decision makers (legislators and the general public) know the risk involved, and are they willing to accept it? For consultants, is this codified to protect us from our legal brethren? As Greg Orsolini pointed out, we need to codify or provide more guidance on how to use the simple nonlinear techniques and to focus on system ductility as opposed to member component ductility and strength. Finally, I would ask researchers to help in the field. How do we really design for the long-duration effects? Are the attenuation relationships the same in these long-duration earthquakes? Are there frequency shifts due to more rapid attenuation of the higher frequencies? These are some future points that we critically need some help with. Thank you.
OCR for page 218
Practical Lessons from the Loma Prieta Earthquake: Report from a Symposium Sponsored by the Geotechnical Board and the Board on Natural Disasters of the National Research Council This page in the original is blank.
Representative terms from entire chapter: