REPORT NO. UCB/EERC-94/01 JANUARY 1994 EARTHQUAKE ENGINEERING RESEARCH CENTER PRELIMINARY REPORT ON THE SEISMOLOGICAL AND ENGINEERING ASPECTS OF THE JANUARY 17, 1994 NORTHRIDGE EARTHQUAKE Preliminary finding from field investigations by teams from the University of California at Berkeley immediately following the earthquake. COLLEGE OF ENGINEERING UNIVERSITY OF CALIFORNIA AT BERKELEY For sale by the National Technical Information Service, U.S. Department of Commerce, Spring­ field, Virginia 22161 See back of report for up to date listing of EERC reports. DISCLAIMER Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the Sponsors or the Earth­ quake Engineering Research Center, University of California at Berkeley. A Report Funded by the f\Jational Science Foundation PRELIMINARY REPORT ON THE SEISMOLOGICAL AND ENGINEERING ASPECTS OF THE JANUARY 17, 1994 NORTHRIDGE EARTHQUAKE Jonathan Bray Gregory Fenves FilipFilippou Stephen Mahin Thomas McEvilly Suzanna Loper Richard McKenzie Siddiq Akbar Douglas Dreger Rakesh Goel Graham Archer Mark Aschheim Scott Ashford Anthony Augello Scott Campbell Susan Chang Chih-Cheng Chin Peter Clark Margaret Ennis William Gookin by Jack P. Moehle, Editor and Faculty Participants Research Staff Michael Riemer Barbara Romanowicz Raymond Seed Nicholas Sitar Christopher Thewalt Andrew Whittaker Patrick Williams Post-Doctoral Research Associates Helen Goldsworthy Guillermo Santana Graduate Student Researchers Nicholas Gregor Carlos Lazarte Dawn Lehman Laura Lowes Abraham Lynn Silvia Mazzoni Kurt McMullin Mike McRae Scott Merry Diane Murbach Michael Pasyanos Gretchen Rau Alvin M. Rodriguez Kenichi Soga Enrico Spacone Jonathan Stewart Bozidar Stojadinovic Patricia Thomas Jorge Zomberg Report No. UCBIEERC-94/01 Earthquake Engineering Research Center College of Engineering University of California at Berkeley January 24, 1994 santana Highlight Editing services provided by Carol Cameron. Copies of this report are available by sending a $15 check made payable to the "UC Regents" to the following address: EERC Reports Earthquake Engineering Research Center 1301 South 46th Street Richmond, California U.S.A. 94804 TABLE OF CONTENTS Acknowledgments Chapter 1: Introduction Chapter 2: Seismological and Geological Observations Chapter 3: Strong Ground Motion Chapter 4: Geotechnical Considerations Chapter 5: Transportation Structures Chapter 6: Building Structures i 1-1 2-1 3-1 4-1 5-1 6-1 ACKNOWLEDGEMENTS This report was made possible by the volunteer efforts of numerous individuals associated with the University of California at Berkeley, as well as other individuals and organizations. The Earthquake Engineering Research Center, under the direction of Professor Jack P. Moehle, organized a reconnaissance team comprising Professors Jonathan Bray, Gregory Fenves, Filip Filippou, Raymond Seed, Nicholas Sitar, Christopher Thewalt, and Jack Moehle; post-doctoral research associates Rakesh Goel, Siddiq Akbar and Guillermo Santana (University of Costa Rica) and Helen Goldsworthy (University of Melbourne, Australia); and graduate student researchers Graham Archer, Mark Aschheim, Scott Ashford, Anthony Augello, Scott Campbell, Susan Chang, Chih-Cheng Chin, Margaret Ennis, William Gookin, Carlos Lazarte, Dawn Lehman, Laura Lowes, Abraham Lynn, Silvia Mazzoni, Kurt McMullin, Mike McRae, Scott Merry, Diane Murbach (of San Diego State University), Gretchen Rau, Alvin M. Rodriguez, Kenichi Soga, Enrico Spacone, Jonathan Stewart, Bozidar Stojadinovic, Patricia Thomas and Jorge Zomberg. EERC staff provided support services. The U.C. Berkeley Seismographic Station organized seismological data under the direction of Professor Barbara Romanowicz. Seismographic Station faculty, staff and students participated in the preliminary analysis, particularly so Dr. Douglas Dreger, graduate students Michael Pasyanos and Nicholas Gregor, and staff members Suzanna Loper and Richard McKenzie, with professors Romanowicz and Thomas McEvilly. Dr. Patrick Williams, of the Lawrence Berkeley Laboratory Division of Earth Sciences, organized geological data. Dr. Williams also organized and led a helicopter reconnaissance of the surface rupture area. Particular credit goes to Professor McEvilly for suggesting this multidisciplinary trip, which also included professors Jack Moehle and Barbara Romanowicz. The United States Coast Guard provided a helicopter and flight crew for the aerial reconnaissance. They went out of their way to be helpful in providing this valuable and highly professional service. Mr. John Williams and Mr. Rick Hughes piloted the helicopter, and AD3 Rick Hughes was flight mechanic. The initial investigative efforts of the geotechnical engineering team were supported, in large part, by the U.S. National Science Foundation. The investigation was aided by several individuals, including Phil Gillibrand of the P.W. Gillibrand Company, Ed Kavazanjian of Geosyntec Consultants, Dean Wise of Browning-Ferris Industries, Deean Affeldt of the PRA Group, Doug Corcorgn of Waste Management, and Mike Courtemarc of Metropolitan Water District. The State of California Office of Emergency Services and the Earthquake Engineering Research Institute established temporary headquarters in Pasadena, which served as a base for some of the reconnaissance activities described in this report. The cooperation and assistance of these organizations and their teams were invaluable. n santana Highlight The California Department of Transportation (Caltrans) authorized complete access to State highway facilities. James E. Roberts, Interim Deputy Director, extended assistance as did the Division of Structures. The Office of Strong Motion Studies, of the California Division of Mines and Geology, under the direction of Anthony Shakal, provided rapid access to strong-motion records that were invaluable for this report. Additional information on the records can be obtained from CDMG/OSMS. Chapter 1 was written by J. Moehle. B. Romanowicz and P. Williams prepared Chapter 2 with contributing authors D. Dreger, M. Pasyanos, S. Loper, N. Gregor, R. McKenzie, T. McEvilly. Chapter 3 was written by S. Chang, G. Santana, J. Bray and R. B. Seed. J. D. Bray, M. Riemer, R. B. Seed, N. Sitar, and J. Stewart prepared Chapter 4 with contributing authors: S. Ashford, A. Augello, S. Chang, C.-C. Chin, M. Ennis, W. Gookin, C. Lazarte, M. McRae, S. Merry, D. Murbach, G. Rau, K. Soga, P. Thomas, and J. Zomberg. G. Fenves and C. Thevvalt prepared Chapter 5 with contributing authors G. Archer, M. Aschheim, S. Campbell, P. Clark, F. Filippou, R. Goel, H. Goldsworthy, D. Lehman, A. Lynn, S. Mazzoni, K. McMullin, J. Moehle, A. Rodriguez, G. Santana, E. Spacone, and B. Stojadinovic. F. Filippou prepared Chapter 6 with contributing authors D. Lehman, L. Lowes, K. McMullin and P. Clark. This report was published with funds from the University of California Earthquake Engineering Research Center. The professional services of Carol Cameron and Katherine Frohmberg at EERC enabled the report to be printed and distributed rapidly. iii santana Highlight santana Highlight santana Highlight santana Highlight PREFACE This report on the seismological and engineering aspects of the 17 January, 1994, Northridge earthquake was printed on 24 January, 1994, one week after the main event. Its purpose is to provide a brief overview of preliminary observations related to the earthquake. The primary audience is seismologists, engineers and related professionals in the earthquake hazard and earthquake risk mitigation field. The report is preliminary in the sense that significant data collection and analysis remain to be completed. Reports containing more complete data and analysis may be issued at a later date. ABSTRACT Immediately following the 17 January, 1994, Northridge earthquake, the Earthquake Engineering Research Center dispatched a reconnaissance team to the epicentral region. This report, issued one week after the earthquake, provides an overview of the seismological and engineering aspects of the earthquake and associated aftershocks. SLIDE SET A slide set containing approximately 1 00 slides obtained during the reconnaissance, including all slides and photographs in this report, is being prepared. Copies of the set are available at cost. To obtain a set, write to EERC Reports, 1301 S. 46th Street, Richmond, California 94804, e-mail to reports@eerc.berkeley.edu, call510-231-9468, or fax 510-231-9461. lV v CHAPTER 1 INTRODUCTION The Northridge earthquake struck the San Fernando Valley region of Southern California at 4:30a.m. local time, Monday, 17 January 1994. The Seismographic Stations at the University of California at Berkeley assessed the main event at moment magnitude 6. 7. According to current accounts, the earthquake resulted in at least 55 deaths and as many as 5000 injuries. The Red Cross estimates 25,000 dwellings are uninhabitable. Very preliminary damage estimates range from $15-30 billion, which, if correct, would make this the most costly natural disaster in U.S. history. Studies of aftershocks and permanent ground deformations are providing data from which will emerge a clear image of the earthquake mechanism and related geological phenomena. Early evidence suggests that the earthquake had a focal depth of about 14 km and a thrust mechanism. The epicenter is approximately 25 km southwest of the epicenter for 1:he 1971 San Fernando earthquake. Ongoing analytical and field work will clarify details of 1:he mechanism. Ground motion records already have been made available from several sources. Durations of strong shaking (peak accelerations exceeding 0.05g) are about 20 seconds in many locations. Several records indicate peak vertical accelerations equal to or exceeding peak horizontal accelerations. Early and approximate analyses of the records suggest that 1:he ground motion intensities may exceed levels commonly used in current engineering design. Preliminary assessments of engineered structures indicate that the majority performed well during the earthquake; however, there is significant and costly damage over a wide geographic region. In most cases the damage appears to have occurred in older structures, 1:he proportions and details of which do not satisfy current requirements for construction. In other cases, damage has occurred in more recent construction. The efficacy of seismic retrofitting and of technologies such as seismic isolation is often evident. Though a significant ~ount: of data has been gathered, the full impact of the earthquake on structural and nonstructural systems will only be understood many months into the future. Immediately following the earthquake, a research team comprising about 50 individuals from the Earthquake Engineering Research Center, Seismographic Stations, and Lawrence Berkeley Laboratories pooled their energies and talents to gather perishable and valuable data on the earthquake and its engineering effects. The team focused its attention on seismology, geology, geotechnical engineering, and structural engineering (transportation and building structures). The five remaining chapters in this report provide brief and preliminary summaries of our findings at the end of one week following the main shock. More detailed summaries and analyses will be made available later. 1-1 Although the Northridge earthquake and its effects have been tragic in the total loss of life, personal injury, and economic losses, we must use this time to advance our knowledge and construction practices. The acceptable performance of the majority of constructed facilities, and the comparatively small number of deaths compared with earthquakes of similar magnitude elsewhere in the world, emphasizes the overwhelming success of several earthquake risk reduction efforts at the national, state, and local levels. It is imperative that these programs continue and expand so that the tragedy of future earthquakes will be reduced. 1-2 CHAPTER2 SEISMOLOGICAL AND GEOLOGICAL OBSERVATIONS This is a preliminary report on the geological and seismological aspects of the January 17, 1994 Northridge earthquake, which occurred at 4:30 am (PST) under the north-western end of the San Fernando Valley, Los Angeles, CA (epicentral location: 34013' North, 1180 32' W, from Caltech). This report is based on main shock and aftershock data from Caltech/USGS, information from analysis of broadband and strong motion records available to UC Berkeley Seismographic Station scientists during the first 5 days following the main shock, as well as geological information obtained in aerial reconnaissance and field investigation conducted by UC Berkeley and Caltech. Seismological observations The results of UC Berkeley's preliminary modelling of broadband records for the main shock from the TERRAscope network and the Berkeley Digital Seismic Network (BDSN) indicate a moment magnitude of 6.7 (local magnitude 6.6 by Caltech), focal depth of -14 km and a thrust mechanism, with both planes striking approximately 100 North of West and dipping approximately 450 (Fig. 2.1). This is consistent with the first motion mechanism released by Caltech several hours after the event. Preliminary results of the empirical Green's function deconvolution analysis in which the effects of source radiation pattern, regional wave propagation, local site conditions and attenuation are removed from the mainshock records, reveals a source duration of approximately 6 seconds (Fig. 2.2). There appears to be a slight directivity towards the North indicating that the event ruptured updip, towards the north, along a south dipping fault. The distribution of aftershocks, covers an area roughly 30km wide (San Fernando to Santa Suzanna) by 25 km long (North Ridge to Santa Clarita Valley) primarily North of the mainshock epicenter, with shallowing depth towards the North (Caltech solutions). The actual fault plane thus appears to be the south dipping plane. The aftershock frequency distribution appears to be consistent with the general trends in California. Several aftershocks of magnitude larger than 5 occurred during the first 5 days after the main shock (Table 2.1). The largest one in that time period occurred at 3:33PM PST on January 17 and has a preliminary moment magnitude of 6.0 (UC Berkeley; Harvard gives 5.9) and a similar mechanism to that of the main shock (Fig. 2.1), with a depth df- 8 km. Reliable moment tensor solutions for some of the largest aftershocks have been obtained at UC Berkeley using body waveform modelling and, independently, surface wave spectral domain inversion. Most indicate thrust mechanisms similar to that of the main shock, although some have slightly rotated strikes towards North of West. There are several strike-slip mechanisms in the center of the aftershock zone (Fig. 2.1). 2.1 Preliminary analysis of strong motion records from TERRAscope and BDSN stations indicate a duration of shaking of -25-30 sec and a possibly complicated rupture with at least 2 shocks separated by several sec (Fig. 2.3). The two shocks appear to be also resolvable in the preliminary deconvolution of the broadband source time function (Fig. 2.2). The Northridge earthquake is the latest and so far the largest , in a series of significant earthquakes that have occurred since 1987 in this part of the transverse ranges. The largest of these were the 1987 Whittier Narrows earthquake (Ml=5.9) and the 1991 Sierra Madre earthquake (Ml=5.8). Both these earthquakes occurred to the east of the Northridge epicenter (figure 2.4). All these earthquakes had similar thrust mechanisms. In contrast to the Northridge event however, they occurred on north -dipping planes, as did the San Fernando (Sylmar) earthquake of February 9, 1971 (Ml= 6.4). The San Fernando event occurred at a depth of 13 km (Heaton and Heimberger, 1979; Langston, 1978; Hanks, 1974) on a previously unmapped fault. For this earthquake, evidence of surface rupture wasfound in a zone directly to the East of the surface projection of the Northridge fault plane. The epicenter of the Sylmar earthquake was located about 25 km northeast of the Northridge event. All these earthquakes are expressions of the north-south compressive deformation occurring across the Transverse Ranges of southern California. This deformation results from the convergence across the "big bend" of the San Andreas fault system between Gorman and Cajon Pass. The thrust mechanism of this earthquake may explain the unusually strong shaking experienced in some areas. Field observations We used the aftershock pattern from the southern California seismic network, the mainshock focal mechanism from UC Berkeley, published geological mapping, and field reports from Caltech to plan a helicopter reconnaissance along the surface projection of the Northridge earthquake fault plane. The reconnaissance was flown on Wednesday January 19th with cooperation from the United States Coast Guard. We observed three areas of extensional ground breakage to the south of the surface projection of the Northridge rupture plane (Fig. 2.5). We believe the primary fault plane is manifested by a broad upwarp, and the upward bending of the mountains has resulted in the opening of many extensional fractures. This is reminiscent of the rupture pattern observed in the Lorna Prieta earthquake. Our observations indicate that extensional surface strain is prevalent across a large part of the Santa Susana Mountains (Fig. 2.6 to 2.9). The pattern of faulting from all data indicates that a major south dipping fault system, possibly an eastward extension of the Oak Ridge fault, produced the Northridge earthquake. 2.2 Interstate 5 Route 14/5 bridge failures are within extensional ground failure that we observed. is possible the strain contributed these bridge in moment inversion. Locations are less reliable are subject to change. Magnitude estimates are more robust. 2.3 ~~~ --------------------~----------~------------~--------------------------~ 10km 10mi Explanation -...._ fault triangles on overthrust side ,...,, dashed where uncertain dolled where inferred or covered 5' e locetions of aftershocks Ill locetions of bridge collapses along Routes 5 and 14 \ I ... ' hill front - area of extensional surface fractures '"''""'1111111111uuu ,, \ \l \\ \ \\ \" \\ \\ \ 1111 II IIIII " \ \ \ \ \ \\ // '""'" ltl 0 -&; Santa Monica Mountains ~ -- ' - ,, Figure 2.1: Map of key seismological features in the area of the Northridge earthquake. The focal mechanism of mainshock indicates pure thrust faulting. Aftershocks indicate that the south-dipping focal plane, is the rupture plane. This plane projects to the surface near the northern edge of the Santa Susana Mountains (where the "Newhall fault" is provisionally located). Aftershocks approximate the location of the Northridge rupture plane. Extensional surface fracturing was documented in the areas marked with hatures. Bridge failures within the zones of extension are located. Note that the 1971 San Fernando (Sylmar) earthquake ruptured the adjacent north-dipping San Fernando fault. 2.4 Source Time Functions for 9401171231 Northridge Earthquake Estimated from Empirical Green's Function Deconvolution bardecon.out 134.48 Location Map -120 -118 B) WestLon. 0 20 40 80 A) seconds Fig 2.2. a) source time functions obtained by deconvolving the motions of a nearly co­ located aftershock with a focal mechanism similar to that of the mainshock. The deconvolution was performed in the spectral domain and the empirical Green's function spectra were corrected with 1% water-level to minimize instability introduced during the deconvolution process. TERRAscope stations BAR, GSC and SBC reveal 6 second source durations. The duration at BDSN station PKDl is shorter (4.9 s) indicating a component of northward directivity durin~ the earthquake rupture. Assumi:nsz a circular fault, a duration of 6 seconds gives a fault radius of 8.2 km. Considering the seismic moment obtained from inversion of complete waveforms (1.2 lo26 dyrie-cm) and a rigidity of 3 1011 dyne/cm2, the average slip on the fault plane is estimated to be approximately 1.9 meters. b) shows the locations of stations used in the analysis. 2.5 .. .. . ... £ •• -·Joo . .. . .. . . ' . .. .. . . • •• e .. ·-- s Fig. 2.3. Excerpt from National Earthquake Information Center epicenter map for the area north-west of Los Angeles, showing epicentrallocations of recent large events relative to that of the Northridge earthquake. The thick line is the coast The scale is approximately 3 em = 10 miles. 2.6 .·- Strong ground motion accelerograms recorded at at u ..... :!r+ ... January 17. 1994 Northridge earthquake traces are radial accelerations were 3.70 (transverse), 2.61 2.7 distance km). Three bott:om respectively. Peak ground (vertical) and 3.89 cm/s2 (radial). Santa Susana Mts San Fernando Valley ..._" folds within Santa Susana Mts -14km Figure 2.5: Geological interpretation of preliminary earthquake source data. A fault plane dipping about 45° to the south is inferred from the focal mechanism and aftershock data. The existence of the Oak Ridge "Newhall fault" is inferred from the seismological data and from preliminary field data. Note that extension at the surface can be produced by the transfer of slip into broad folding near the tip of a thrust 2.8 l'Y 0 \0 Figure 2.6: Ground ruptures within the Newhall-Potrero Oil Field along the northern edge of the Santa Susana Mountains. ' ~ Figure 2.8: Displacements along steeply souUl-dipping bedding planes to the east of the Newhall-Potrero oilfield. Figure 2.7: Ground ruptures within the Newhall-Potrero Oil Field along the northern edge of the Santa Susana Mountains. Figure 2.9: Typical rockfall, northwest of the Route 14/5 interchange, Santa Susana Mountains. CHAPTERJ STRONG GROUND MOTION The Northridge Earthquake of January 17, 1994 generated a large number of strong motion recordings over a wide variety of geologic site conditions, including free-field stations on rock and soil as well as recordings of motions from instrumented structures of varying types of construction. Several agencies, such as the California Division of Mines and Geology (CDMG) Strong Motion Instrumentation Program (CSMIP), the U.S. Geological Survey (USGS), and California Institute of Technology (Caltecb) each maintain relatively extensive strong-motion instrumentation networks in the affected region. As of January 21, 1994, the only strong motion records that have been preliminarily processed and made publicly available are those from 44 instrumentation stations of the CSMIP network (1994). Refer to Table 3.1. Figure 3.1 is a map of the epicentral region showing the locations of selected CSMIP stations. Thirty-eight out of the 44 available accelerograms from Strong Motion Instrumentation Program (CSMIP) stations were analyzed, as information regarding the site geology at six of the sites is not yet available. Table 3.2 presents a brief summary overview of the general site geology at each of the 38 stations based on data provided by CSMIP. Figure 3.2 is a plot of peak horizontal ground acceleration vs. epicentral distance, for both free-field records and records obtained at the bases of structures. These are further separated by use of different symbols for records obtained at stations sited "on soil" or "on rock". It should be noted that epicentral distance is a generally poor measure of "distance", especially in the near-field, and that closest distance to the fault rupture surface is generally to be preferred. Unfortunately, there continues to be debate regarding the precise location of the rupture surface, so epicentral distance bas been used herein. As shown in Figure 3.2, all recorded peak horizontal accelerations from the free-field rock sites plotted above the mean attenuation relationship for rock as proposed by Joyner and Boore (1988). The "closest" (based on epicentral distance) free-field instrument on rock, located at Pacoima-Kagel Canyon Fire Station #74 approximately 17 km northeast of the epicenter, recorded a peak ~orizontal acceleration of 0.44g. The largest free-field peak acceleration recorded on rock was 0.49g, recorded at the Los Angeles 7-story University Hospital, 36 km southeast of the epicenter. The closest free~ field instrument on soil (approximately 10m of alluvium over siltstone) is located at Tarzana-Cedar Hill Nursery, approximately 7 km south of the epicenter. Peak horizontal and vertical accelerations of 1.82g and 1.18g, respectively were recorded. It should be noted that the Tarzana station recorded much higher accelerations than stations with similar epicentral distances during the 1994 Northridge earthquake, as well >1 and 1991 Sierra Madre Earthquakes. However, recordings from this station during the 1992 Big Bear and 1992 Landers Earthquakes show reasonable accelerations for this site. Peak horizontal ground accelerations from strong-motion instruments located at the bases of structures are also shown in Figure 3.2. It is likely that these values are, on the average, slightly lower than what would be recorded at a free-field site since at sites where both free-field and ground/basement floor recordings were available, the peak horizontal accelerations recorded at the structure sites were generally lower then the free-field measurements by approximately 10 to 30 percent Unlike the 1989 Lorna Prieta Earthquake, no clear trends in amplification of ground motions at soil sites is apparent for the initial 38 stations studied in this report; however, a preliminary map of heavily damaged (unsafe) buildings prepared by the City of Los Angeles Department of Building Safety shows clusters ofdamage concentrated in east-west trending zones along Interstate 10 between Santa Monica and east Los Angeles, through Hollywood between Interstate 101 and Interstate 5, along Highway 134 east of Interstate 405 in She1man Oaks, as well as in the epicentral area. Fmther investigation of site geology, structural basin effects, seismological and structural considerations will be required to determine how local site conditions may have contributed to these significant clusters of damage. Figure 3.3 presents a plot of peak horizontal accelerations recorded at the 38 CSMIP stations vs. the peak vertical accelerations recorded at these stations. Although at a few stations vertical accelerations recorded were nearly equal to the recorded horizontal accelerations, and at one station the peak vertical acceleration was higher than the peak horizontal acceleration, in general, peak vertical accelerations were more typically equal to approximately two-thirds of the peak horizontal accelerations. Table 3.3 summruizes the results of preliminary analysis of selected records obtained in and near to the epicentral region. The predominant period was read directly, due to the fact that no digitized records are available yet. The records used for this purpose are only the five nearest to the epicenter. They include both the Tarzana and the Sylmar records, which have shown unusually high values of peak acceleration. As shown in Table 3.3 the records studied show a predominant period of about .25 to .4 seconds. Also, when considering the level of acceleration above 0.05g as indicative of the duration of the stronger phase of shaking, it appears that the duration of strong shaking was on the order of 15 to 20 seconds at and neru· the epicentral region. These preliminary results indicate that the destructive potential of this eruthquake was somewhat higher than the levels observed in the urban ru·eas of Northern California during the Lorna Prieta earthquake of 1989, as both higher near-field accelerations and a slightly longer duration of strong shaking appear to have been produced by the Nmthridge Ea1thquake fault rupture. 3-2 Finally, Figure 3.4 shows plots of the free-field acceleration time histories recorded at (a) the Tarzana-Cedar Hill Nursery (CSMIP Station #24436) and (b) the Sylmar County Hospital Parking Lot (CSMIP Station #24514). These plots are taken directly from the CSMIP preliminary reports, and are poorly reproduced. Nonetheless, they serve to illustrate the character of the motions at these sites, which are notable for their considerable duration of relatively strong shaking. The Cedar Hill Nursery site was the station that recorded the highest "level ground" accelerations released to date, and the County Hospital station is of particular interest as it is adjacent to the site of the Olive View Hospital which fared poorly in the previous (1971) San Fernando Earthquake. 3-3 Table 3.1 :Data Recovered from Selected Stations of the California Strong Motion Instrumen­ tation Program (CSMIP) for the 17 January 1994 Northridge/San Fernando Valley Earthquake Station Coordinates Epicentral Maximum Acceleration No. Station Name N.Lat W.Long Distance Free-field Base Struct. 24386 Van Nuys- 34.221 118.471 6km --- 0.47g H 0.59g H 7-story Hotel 0.30g v 24436 Tarzana- 34.160 118.534 7km 1.82g H --- --- Cedar Hill Nursery U8gV 24087 Arleta- 34.236 118.439 9km 0.35g H --- --- Nordhoff Ave. Fire Station 0.59g v 24322 Sherman Oaks - 34.154 ll8.465 10km --- 0.46g H 0.90g H 13-story Commercial Bldg. 0.18g v 24514 Sylmar- 34.326 118.444 15 km 0.91g H 0.82g H 2.3lg H 6-story County Hospital 0.60g v 0.34g v 24088 Pacoima- ":\4.288 118.375 l7km 0.44gH --- --- Kagel Canyon Fire Sta. #74 0.19g v 24207 Pacoima Dam 34.334 118.396 18 km --- 0.54g H >2.3g H 0.43g v >1.7g v 24464 North Hollywood - 34.138 118.359 19 km --- 0.33g H 0.66gH 20-story Hotel 0.15g v 24231 Los Angeles - 34.069 118.442 19 km --- 0.29g H 0.77g H 7-story University Bldg. --- 0.25g v 24389 Century City - 34.064 118.417 20km 0.27g H --- --- LACC North O.l5g V 24643 Los Angeles - 34.059 118.416 21km --- 0.32g H 0.65gH 19-story Office Bldg. 0.13g v 24385 Burbank- 34.187 118.311 2lkm --- 0.30g H 0.79g H 10-story Residential Bldg. 0.13g v 24370 Burbank- ' 34.185 118.308 22km --- 0.35g H 0.49gH 6-story Commercial Bldg. 0.15g v 24670 Los Angeles - 34.031 118.433 23 km --- --- l.OOg H Il 0/405 Interchange Bridge 1.83g v 24303 Los Angeles - 34.090 118.339 23 km 0.41g H --- --- Hollywood Storage Bldg. Free Field 0.19g v 24236 Los Angeles - 34.090 118.338 23 km 0.41g H 0.29g H 1.61g H Hollywood Storage Bldg. 0.19g v O.llg V 24538 Santa Monica - 34.011 118.490 24km 0.93g H --- --- City Hall Grounds 0.25g v 24251 Wood Ranch Dam 34.240 118.820 26km --- --- 0.39g H 0.18g v 24157 LA - Baldwin Hills 34.009 118.361 28km 0.24g H --- --- O.lOg V 24612 Los Angeles - 34.043 118.271 31 km 0.19g H --- --- Pico and Sentous 0.07g v 24602 Los Angeles - 34.051 118.259 32km --- 0.15g H 0.4lg H 52-story Office Bldg. O.llg V 24611 Los Angeles - 34.059 118.246 32km 0.19g H --- --- Temple and Hope O.lOg V 24655 Los Angeles - 34.021 118.289 32km --- 0.29g H 1.2lg H 6-story Parking Structure 0.22g v 0.52gV 24629 Los Angeles - 34.048 118.260 32km --- 0.14g H 0.19g H 54-story Office Bldg. 0.08g v 24652 Los Angeles - 34.021 118.287 32km --- 0.24g H 0.59g H 6-story Office Building 0.08g v 0.18g v 3-4 I . . I Station Coordinates I Epicentra! i Maximum Acceleration . -l . , Station Name N.Lat l W.Long--j Distance* 11 Free-field 11 Base Str.Jct~-~ 69.4·r1 .. ~==~~ =======~~~~~==~~~=-4·==~7===F· ====~~==~·=7~~~==~~-=-~~~ U:ls Angeles- 34.058 I 118.249 32 km I ·· · 0.21g H ·029g Hi ! J 15-story Govt. Office Bldg. I' I 0.07g V I 1 24579 j Los Angeles- 34.044 118.261 32 km O.l8g H 0.34g H I • 1 9-story Office Bldg. ~~ .. I· 0.12g V I 24283 I Moorpark 34.288 118.881 33 km 0.30g H I I O.l5g v I ! 14654 1!1.' E! Segundo - 33.920 118.390 36 km O.l3g H 14-story Office Building 0.0.-.lg V 24605 U:ls Angeles- 34.062 U1U98 36 km 0.49g H 0.37g H 1 7-story University Hospital 0.12g V 0.09g V i 24541 I 24468 i 24592 I ~~24SW 24401 14606 14406 14560 14533 14578 23622 23631 23631 (Base lsolated) Pasadena- 6-story Office BuiJding Los Angeles - 8-story CSULA Admin. Bldg. U:ls Angeles - City Terrace U:ls Angeles - Fire Command Control Bldg. (Base Isolated} San Marino- Southwestern Academy Whittier- 8-story Hotel U:ls Angeles - Vincent Thomas Bridge Long Beach City Hall Grounds U:lng Beach .. 15-story Govt. Office Bldg. Seal Beach- 8-story Office Bldg. (Base Isolated) San Bernardino ·· 1-story Commercial Bldg. San Bernardino - Hwy H0/215 Free Field San Bernardino - 34.146 118.147 34.067 118.168 34.053 H8.17l 34.053 118.171 34.115 H8J30 33.975 118.036 33.7:50 118.271 33.768 118.196 33.768 H8.195 I 33.757 118.084 34.098 117.293 34.065 117.292 1!2636 H0/215 Interchange Sage- Fire Station '-·- ..J_ _________ _ 34.064 H7.296 I' i 33.580 l !6.931 _L ___ .......L. __ _L 37 krn 38 km 39 km 39 km 39 km 54 km 58km 59km 59km 66km 115 km 115 km 115 km 165 km 0.32g H 0.13g v 0.32g H O.l3g V 0.16g H 0.09g v 0.06g H 0.03g v 0.06g H 0.03g v 0.09g H 0.04g v O.lOg H 0.04g v 0.10g H 0.04g v 0.03g H 0.02g v 0.!7g H 0.09g v 0.17g H 0.06g v 0.22g H OJlg V 0.19g H O.lOg V 0.25g H 0.08g v 0.04g H 0.03g v 0.08g H 0.03g v 0,05g H 0.02g v OJ3gH 0.04g v 0.25g H O.l7g V 0.2lg H 0.13g v 0.2lg H 0.25g H O.l7g V 0.35g H 0.30g v 0.49g H 0.65g H 0.44g v 0.06g H 0.05g v 0.15g H O.l6gV O.l5g H 0.47g H 0.3lg v Table 3.2: Site Geology at Selected CSMIP Stations -·--·-·--· No. Station Name Site Geology No. Station Name Site Geology 24386 Van Nuys- 7-story Hotel Alluvium 24629 LA - 54-story Office Building Alluvium over sedimentary rock 24436 Tarzana- Cedar Hill Nursery Shallow alluvium (-10m) over 24569 LA - 15-story Government Office Bldg. Siltstone siltstone 24579 LA- 9-story Office Building Alluvium 24007 Arleta - Nordhoff Ave. Fire Station Deep alluvium 24283 Moorpark Alluvium 24322 Sherman Oaks-13-story Commercial Bldg. Alluvium 24605 LA-7-story Univ. Hospital (Base Isolated) Siltstone 24514 Sylmar - 6-story County Hospital Alluvium 24541 Pasadena - 6-story Office Building Deep alluvial fan 24088 Pacoima- Kagel Canyon Fire Station #74 Sandstone 24468 LA - 8-story CSULA Admin. Bldg. Siltstone 24207 Pacoima Dam Metamorphic dioritie gneiss 24592 LA - City Terrace Siltstone 24464 North Hollywood - 20-story Hotel Sandstone/shale 24580 LA - Fire Command Control Bldg. (Base Siltstone 24231 LA- 7-story University Building Terrace deposits Isolated) 24389 Century City - LACC North Terrace deposits 24401 San Marino - Southwestern Academy Deep alluvial fan 24385 Burbank - 10-story Residential Bldg. Alluvium 14606 Whittier - 8-story Hotel Shallow alluvium over sedimentary bedrock 24370 Burbank - 6-story Commercial Bldg. Alluvium 14406 LA - Vincent Thomas Bridge Alluvium 24303 LA - Hollywood Storage Bldg. Free Field 130m alluvium 14560 Long Beach - Oty Hall Grounds Terrace deposits 24236 LA - Hollywood Storage Bldg. 130m alluvum 14533 Long Beach- 15-story Govt. Office Bldg. Terrace deposits 24538 Santa Monica - City Hall Grounds Terrace deposits 14578 Seal Beach • 8-story Office Bldg (Base Alluvium 24157 LA - Baldwin Hills lm fill over shale/sandstone Isolated) 24612 LA - Pico and Sentous Alluvium 23622 San Bernardino - 1-story Comm. Bldg. Deep alluvium 24602 LA - 52-story Office Building 7m alluvium over sedimentary 23631 San Bernardino- Hwy 110/215 Free Field Alluvium rock 23631 San Bemardino-Hwy 110/215 Interchange Alluvium 24611 lA · Temple and Hope Siltstone 12636 Sage - Fire Station Shallow alluvium over granitic bedrock 3-6 Table 3.3: Preliminary Summary of Data for 5 CSMIP Stations At and Near To the Epicentral Region Station Site Epicentral Predominent Duration Characteristic Name Condition Distance (km) Period (sec) (sec) Frequency (Hz) Sylmar (E-W) Alluvium 15 0.35 14 5.71 Arleta (E~ W) Deep Alluvium 9 0.40 16.5 4.55 Tarzana (E-W) Alluvium (lOrn?) 7 0.35 20.5 7.80 over siltstone LA Storage (N-S) Alluvium (130m?) 23 0.24 15 6.40 over sandstone shale LA Pico (N-S) Alluvium 31 0.40 13 4.15 "'Duration of aa:eleratioiiS greater tiwl 0.05g. 3-7 ~ g ·--Cll3 b '1l u u < i g .~ ... C) = i G.ll ~ 10.000 0 fl. D 1.000 <> 0.100 0.010 Free-field Rock Site Free-field Soil Site Structure Base on Rock Structure Base on Soil Mean+ 2 Standard Deviation. Mean Mean - 2 Standard DevJatlon 0.001 L----'-----'1..-...JL......I..-L-JU-.I..I.---'-........JI..-...Jb...L....L-I....I...L.L---'-........JI..-...JL......L....L-J...L.J....I 1 10 100 1000 Distance (km) Figure 3.2: Peak Horizontal Acceleration w. Epicentral Distance, and the Joyner and Boore (1988) Attenuation Relationship. 3-9 2.00 " " " 0 Free-field Rock Site " " ll. Free-field Soil Site " " 1.60 0 Structure Base on Rock " " - Structure Base on Soil " 01) ~ " -a " 0 " ... " .. 1.20 " ... / ~ "iS / ~ " ~ < " " -; " ~ " ... - 0.80 t " > " " ...:.c " ! ll. / ll. " " " 0 0.40 " ~ 0.00 ~-=--'-----L-----L--.I...--....l.---L----L--.I...---'-----1 0.00 0.40 0.80 1.20 1.60 2.00 Peak Horizontal Acceleration (&) Figure 3.3: Plot of Peak Horizontal Acceleration vs. Peak Vertical Acceleration at 38 CSMIP Strong Motion Recording Stations 3-10 "f ..... ..... Tarzana - Cedar Hill Nursery . (CSMIP Stat ion 24436) - ..... --- -~~- - ......... .........._. ............_,...,....... - - .. 1 a .... ..J;. - Record 24436-51614-94017 --- ---- --- ~-Max. - - - - - - -- '1~fel. go~~- L82 - ill 11 . 11 Jf .I 1;1'1 Unlh .. l1~ ~Ill I Hit II NV I I MIIII~Af I ¥ 11 ' ' ' • Up 1.18 360°~ 1.01 4 15 20 Sec. Sylmar - 6-story County Hospital Parking Lot (CSMIP Station 24514) Record 24514-55254-94017 .... .,.-.. ..... ~-..~~--.---..~--.--.-.--..-_~~-.-.~·~~~el. .• ~J 356ft' ~'tt~ ~ ~ ~, ,1171 err •o.6t --Gm------~~-.~~~--~---.----~-------------------------------------------------------------Up 0.60 esc- 0.91 .\...\-'-'-.,_ ,_ ~ ,_ .,_ .,_ '-'-'-'-'-,_ '-'-,_ '-'-'-'-'-'-'-'-'-'-'-'-'-,_, ,_ ,_ ,_ ,_ ,_ '-'\. ,_ '-\.. ,_ .__. 0 1 2 3 A 5 10 15 20 Sec . Figure 3.4~ Free-Field Acceleration Time Histories Recorded at CSMIP Stations #24436 and 24514 (CSMIP, 1994) CHAPTER4 GEOTECHNICAL CONSIDERATIONS The Northridge earthquake of January 17, 1994 caused extensive damage throughout 'the epicentral region and in several surrounding areas. Based on preliminary reconnaissance work performed immediately following the earthquake, the major geotechnical aspects of this event were found to include the following: 1. Pronounced ground movements were observed at Potrero Canyon, southwest of Magic Mountain. The observed movements mainly consisted of extensional features, but localized compressional features were also found. This ground breakage most likely resulted from areal subsidence and lateral spreading at the bedrock-alluvium contact; but the widespread surface distress in the regions overlying the northern edges of the apparent shallow southernly dipping thrust fault rupture plane may have been due in part to fracturing and distress within the folded upthrown bedrock adjacent to the primary thrust fault. 2. Local site conditions do not appear to have exerted as dominant an influence on ground shaking levels as in the recent 1989 Lorna Prieta Earthquake, and the largest concentration of structural damages appears to have occurred in the heart of the epicentral region at and near Northridge. Several significant concentrations of damage occurred away from this epicentral region, however, (a) at Hollywood, north of Santa Monica Blvd. and between Highways 5 and 101, (b) at Sherman Oaks, near Highway 101 just east of Highway 405, and (c) along an arc in central Los Angeles just to the northeast of Culver City. All three of these regions of concentrated "clusters" of structural damage appear to be underlain by pronounced alluvial basins. The effects of deeper, structural basins on ground motions may also be significant, as they appeared to be in this same region in the previous 1971 San Fernando Earthquake, but there are not yet enough strong motion records available to properly investigate this. 3. Soil liquefaction and lateral spreading occurred over large areas in Northridge, near the junctures of Highways 5 and 210 and 5 and 405, and in Simi Valley. Although of relatively large overall areal extent, liquefaction and lateral spreading were of minor severity throughout most of these regions, and appeared to contribute little to the structural damages that resulted primarily from strong shaking and inertial forces in these regions. Liquefaction and lateral spreading (and compression) was evidenced mainly by curb and pavement damages, and resulted in numerous small pipe breaks in these regions. More pronounced liquefaction occurred at and near the Jensen Filtration Plant near Upper Van Norman Lake, but damage to the facility itself~ relatively minor. The Juvenile Hall landslide, which occurred as a result of liquefaction near the juncture of Highways 5 and 405 during the 1971 San Fernando Earthquake, experienced minor downslope movements, offsetting curbs by approximately 3 to 6 inches or less along the approximate boundaries of the 1971 slide zone. Signs of liquefaction (minor lateral spreading, compression and/or sand boils) were also detected at various sites up to 27 miles from the epicenter, including 4-1 along the Santa Clara River between Fillmore and Highway 5, in Potrero Canyon, in the dry lakebed behind Hansen Dam, in Santa Monica, and on the coast at Marina del Ray, at Kings Harbor at Redondo Beach, and at the western end of the Port of Los Angeles. 4. Numerous landslides and rockfalls occurred near the coast at Pacific Palisades and in sparsely populated regions in the Santa Monica Mountains, the San Gabriel Mountains east of the Highway 5 and 14 interchange, and the Santa Susana Mountains. A significant coastal bluff failure occurred at Pacific Palisades, destroying several homes and closing the Pacific Coast Highway (Hwy. 1). Slope movements may also have been instrumental in disabling two aqueduct pipelines, resulting in loss of water service to the Simi Valley region. Overall, however, the widespread occurrence of slope failures, rockfalls and ravelling in natural slopes and talus occurred primarily in undeveloped areas and caused little damage. 5. There are a large number of earth and rockfill dams in the strongly shaken region. A number of these experienced minor deformation and cracking, and minor slope movements occurred in several natural abutment slopes. Several embankments suffered minor damages at and near their crests, and the Pacoima Dam (a concrete dam) suffered damages very similar to those it experienced in the 1972 San Femando Earthquake. The single "failure" of a "dam" was actually the loss of a small dike (approximately 20 feet high) retaininga minor pond at the influent basin of the DWP treatment facility near Van Norman Reservoir. There were, however, no significant occurrences Qf distress at major dams posing significant risk of failure, and overall performance of earth and rockfill dams appears to have been good. 6. Nine major solid waste landfills in the strongly shaken region were inspected. Several of these sustained some minor cracking within their surface cover soils, necessitating some minor re-working and re-compaction of the cover soils to reduce gas leakage (and odor). There were, however, no indications of significant distress to slopes or geosynthetic liner systems, and overall stability and performance of these fills appears to have been very good. One of the major landfills (the Oil Landfill) is well­ instrumented with survey monuments, inclinometers, and a pair of strong motion recording stations (on the crest and adjacent to the toe of the fill). Well-documented seismic performance data for waste landfills is currently very sparse, and the data provided by this event can be expected to be of major value to designers of -waste landfills. 7. Numerous small pipe breaks occurred in areas affected by liquefaction and/or Dlinor lateral spreading, including Northridge and the greater western San Fernando Valley area, and Simi Valley. Damage to two major aqueduct pipelines resulted in prolonged loss of water service in the Simi Valley area. A rupture in an oil pipeline resulted in significant contamination of the Santa Clara River at and west of Highway 5. Overall, performance of water systems was very good, and water service was restored to most areas by Wednesday evening. Although minor in impact, damages related to geotechnical considerations were widespread, as shown in Figure 4.1. Due to the brief time period between the completion 4-2 of our field reconnaissance and the publication of this report, most photographs of the various features described herein were unavailable for this first report. We regret this inconvenience, but refer the reader to an upcoming report to be published through the EERC which will more thoroughly document the geotechnical aspects of this earthquake. Ground Failure The Northridge Earthquake caused ground failures at several locations within the San Fernando Valley and the Los Angeles basin from Highway 126 in the north to the Port of Los Angeles in the south. Although these phenomena were concentrated mainly in the epicentral region, incidents of ground failure or ground deformation did occur at distances of up to 36 miles from the epicenter. Significant surface breakage occurred on the north flank of the Santa Susana Mountains, south of State Road 126, at Potrero Canyon. The valley, which is approximately 2.5 miles long and less than a mile wide, is covered mainly by landslide debris -and alluvial material. The area was inspected and where surface breakage occurred detailed maps were developed. Unear ground breakage features, which are roughly parallel to the east-west trend of the mountain ridges, were observed along the northern and southern margins of the valleys within Potrero Canyon. The ground fractures tend to follow the topographic contours around the base of the hills, but they do cut linearly across alluvial deposits at numerous locations. On the northern margins of the valleys, the fractures are primarily extensional, with minor right-lateral offsets. Multiple ground fractures within zones 5 to 30 feet wide accommodate as much as 2 feet of vertical movement (see Figure 4.2). The width of the fractures vary from less than ~-inch to as much as 4 inches. Extensional features are also observed along the southern, eastern and western margins of the valleys. Minor left-lateral offsets occur along the southern margins of the valleys. Compressional features, however, are found along the southern margins of the valleys at a number of locations. These features include shallow thrusting along distinct shear surfaces that dip to the south at approximately 30 to 40 degrees with up to 6 inches of dip-slip displacement. Headscarps in the hills above these thrust features, which might have indicated that they represented toes of landslides, could not be found. Evidence of localized compression was also noted at the entrance to valleys (e.g., see Figure 4.3), but a majority of the significant ground fracturing in this area was extensional. This site is being investigated for possible evidence of surface fault rupture, but it appears that much of these previously stated observations can be explained by earthquake shaking-induced compaction of alluvial sediments and large-scale lateral spreading. The surface area of the alluvial deposits bounded by these ground fractures, which are predominantly extensional, is over 2000 acres. The geology in this region is extremely complex and this situation is exacerbated by the fact that oil and water have been withdrawn from the area over the last hundred years. Further investigation, including trenching and surveying, is warranted at this site. 4-3 Soil Liquefaction and Lateral Spreading Evidence of soil liquefaction including sand boils, ground settlement, and lateral spreading was found over a fairly widespread area, as shown in Figure 4.1. Damage associated with liquefaction generally included breakage of buried pipelines and pavement cracking/buckling. Based on the preliminary results of our reconnaissance, liquefaction does not appear to have directly contributed significantly to any structural failures of buildings or highway structures, and was a relatively minor factor as compared to strong shaking in terms of damage to structures in most areas. Much of the "lateral spreading" damage noted in urban areas appears likely to have resulted either from minor liquefaction at depth, with non-liquefied overlying soils largely mitigating surface distress, or from cyclic compaction of non-saturated alluvium. The northernmost area found to show evidence of liquefaction to date was along the Santa Clara River between Highways 23 and 5. A thorough reconnaissance of this area was impossible due to an oil cleanup operation. However, sand boils were found in Potrero Canyon and surrounding areas, and adjacent to bridge piers for the crossing of High'Way 23 over the Santa Clara River. An example of the Potrero Canyon sand boils, which have a significant fines content, is shown in Figure 4.4. Near a particular pier under construction at the bridge for the Highway 23 crossing over the Santa Clara River, sand boils were observed near the pier and cracks induced by lateral spreading were found approximately 15 feet away from the pier .. No significant damage occurred to the bridge structure as a result of this liquefaction. Liquefaction was also observed by a local resident at a site along the Santa Clara River pear Piru. Based on the reports of this individual, sand boils emerged during both the main shock and a magnitude 5.5 aftershock, but not during somewhat smaller 5.1 aftershocks. Having thus bracketed the earthquake magnitudes wherein liquefaction occurred, this site may represent an interesting case history against which to calibrate liquefaction analysis procedures. In the 1971 San Fernando earthquake, the Lower San Fernando dam suffered extensive damage and was nearly overtopped due to sliding of the upstream shell induced by liquefaction of a portion of the shell material. Following the 1971 earthquake, the dam embankment was repaired and the crest lowered, but the repaired embankment was not intended to impound water. In the recent earthquake, this area again experienced liquefaction. Sand boils, sand fissures, and earth fissures were observed approximately 120 to 500 feet from the upstream toe of the now inactive embankment. The earth fissures were oriented parallel to the axis of the dam, were up to 8 inches wide with 8-inch vertical offsets, and were generally 120 to 250 feet from the upstream toe. Earth and sand fissures oriented perpendicular to the axis of the dam were more prevalent in the region beyond 300 feet from the upstream toe. The Jensen Filtration Plant, adjacent to the San Fernando Dam complex, experienced damage from ground movement and lateral spreading which forced a shutdown of the facility. This facility had been heavily damaged as a result of extensive liquefaction in the 1971 San Fernando Earthquake, and the damages in the current (Northridge) event were generally similar but were significantly less severe. Damage at the site included: Earth fissures 200 feet long and up to 3 inches wide with a maximu:rn of 8 4-4 inches of vertical offset in the parking lot of the main control building. This parking lot is located at the top of a 40 foot high slop~, just above the DWP Treatment Facility. • Settlement of the ground surface adjacent to the main control building with a maximum settlement of about 4 inches. • Several pipeline breaks, including the main influent aqueduct, as well as irrigation lines and chlorination lines. • Minor horizontal and vertical movements across construction joints in the pipeline gallery below the main control building and reportedly in the sedimentation basins. Many of these joints had also moved differentially during the 1971 San Fernando Earthquake. • Apparent partial floating of an underground 50 million gallon finished water reservoir, with movement on the order of 2 to 4 inches. The ground movement causing the damage noted above may have resulted in part from liquefaction of the loose alluvial soils underlying nonliquefied fills placed during construction at the treatment facility, as occurred in the 1971 San Fernando Earthquake. No sand boils or other direct evidences of liquefaction were observed on the site; however, sand boils were observed at the base of the slope west of the facility within the DWP treatment facility. A large ground movement, encompassing a portion of the San Fernando Juvenile Hall facility, occurred as a result of soil liquefaction during the 1971 San Fernando earthquake. The slide appears to have been partially reactivated during the 1994 Northridge earthquake, as large fissures were observed in the parking lot at the southeast comer of the facility. These fissures were parallel to a large sewer line, running southwest-northeast below the parking lot, and were up to 4 inches wide with little or no vertical offset. Additional evidence of partial slide reactivation was gathered along San Fernando Road, southwest of the Juvenile Hall facility. Cracking of the pavement and cracking and buckling of curbs were observed along the east side of San Fernando Road in the vicinity of the previously mapped landslide boundary. In addition, ground cracking was observed in a DWP facility on the west side of San Fernando Road which also appeared to correspond with the previously mapped landslide boundary. Farther east, liquefaction was observed in the dry lake bed behind Hansen Dam, a U.S. Army Corps of Engineers flood control dam that at the time of the field inspection was not impounding water. Sand boils up to 3 feet in diameter, and sand fissures up to 50 feet long and 6 inches wide, were observed upstream of the reservoir flood zone across an approximately 300 by 1000-foot area near several ponds. Sand boils from this area are shown on Figures 4.5 and 4.6. Surprisingly large flows were observed to have been exuded from several boils and fissures which resulted in localized erosion. Lateral spreading of up to 3 feet and settlements of about one foot were also noted in some areas. The city of Granada Hills, located north of Northridge, and northern and central Northridge experienced significant ground movement as evidenced by numerous cracks in streets and broken and buckled curbs (Figures 4. 7 and 4.8). Significant ground fracturing was found west of Woodley Ave., south of Midwood Dr., east of Shoshone Avenue, and north of Highway 118. Figure 4.9 is an example of a compressional feature due to lateral spreading in this general area, whereas Figure 4.10 is an example oflateral spreading clearly 4-5 related to soil liquefaction. The cracks are typically up to 2 inches wide in the asphalt concrete pavement, although in one case a 6 inch separation between a house foundation and the adjacent ground was noted. The ground cracking generally trends east to west and continues for more than one mile. This area was also the site of numerous water and gas pipe breaks. The water pipes (68 and 48 inch diameter) and the gas pipe (12 inch diameter) on Balboa Boulevard just north of Rinaldi Street, were separated by 8 to 12 inches according to the repair crew at the site. At one location nearby, a maintenance crew reported 'that a 1-inch diameter pipe separated six inches and was displaced six inches laterally. Ground failure occurred on the slope at the south side of Highway 118 just east of the Balboa Boulevard overpass. Cracks up to 6 inches wide with up to 8 inches of vertical offset were observed on the south side of Highway 118 just east of the overpass. At the eastbound on-ramp to Highway 118 at Balboa, earth fissures subparallel to the highway were found near both sides of the on-ramp. Erosion from a broken water main near the overpass caused a large void to form (exposing several piers) near the south abutment. Many large cracks in the pavement were observed along Highway 126 between Fillmore and Interstate 5. The probable cause of these cracks is settlement and lateral spreading of the fills underlying the roadway. In Fillmore, severe cracking of asphalt pavement and concrete curbs was noted at the intersection of Celis and Wolfskill. A water line also appeared to have been broken in the area. Lateral spreading and settlement occurred at numerous locations throughout the eastern end of the Simi Valley area, causing minor slope displacements and damage to pavements and buried utility lines. Numerous pipeline breaks occurred in this area. These movements may have resulted in part from liquefaction of loose alluvial sands underlying nonliquefied surficial soils; however, no sand boils were observed and the occurrence of liquefaction cannot be confirmed. The most dramatic example of such movements occurred at Rory Lane just north of the Arroyo Simi drainage channel where approximately 8 to 12 inches of lateral and vertical offsets were observed. A large block of material appeared to have displaced southwards towards the channel. Evidence of liquefaction in the form of sand fissures and sand boils was observed in the southern half of the northwest parking lot at the Santa Monica Municipal Pier. Extensive cracking of the 5-inch thick asphalt pavement was typically oriented subparallel to the coastline. Lateral and vertical offsets were generally about 1¥2 inches, although extension cracks of up to 5 inches were also recorded. No signs of liquefaction were observed below the Municipal Pier or on the beach adjacent to the parking lot. Damage due to liquefaction was also observed in the King Harbor area of Redondo Beach. At an artificial (man-made) beach and swimming area south of Portofino Way and Harbor Drive, numerous sand boils of up to 4 feet in diameter were found. Cracks with vertical offsets of up to 2 inches were located concentrically around the swimming area. At Marina del Rey, a large fissure possibly due to liquefaction was reported by a representative of the Department of Harbors and Beaches at an artificial beach between Palawan Way and Panay Way. Representatives of the EERC were unable to inspect the damage, as clean-up of the area occurred prior to their arrival. 4-6 North of this area, extensive damage occurred at Marina Way and Harbor Drive. Although the cause of the damage was not apparent, two sand boils were observed at the site, and the south retaining wall along Marina Way had bulged and displaced southward. Damage in the area consisted of a broken 8-inch sewer line along the center of Marina Way, and cracking and buckling of asphalt pavement along most of the length of Marina Way. Evidence of ground movement was also observed along Nagoya Way on the western side of the Port of Los Angeles. The area of significant ground movement was located between Berths 83 and 76, and possibly extended south to Berth 74. A reinforced concrete bulkhead was located along its entire length. In the parking lots at the north end of Ports 0' Call Village, the ground adjacent to the bulkhead settled as much as 1 to 1 ~ inches. Cracks in the asphalt pavement (between 1/16 and 3/4 inches wide) or new separations between pavement and sidewalk were observed subparallel to the bulkhead, extending through most of the length of the parking lots. Cracks in the concrete slabs-on-grade were also observed within some of the buildings in this area. Cracks subparallel to the bulkhead as wide as Yz inch were observed in the brick-covered walkways between buildings. The worst cracking was observed near Berth 77, which opened up at the top of the bulkhead, cracking the brick patio and the slabs-on-grade inside the structures. Evidence of wall movement was also observed at this location, and a gas pipe broke nearby. Adjacent to and north of Berth 83, the paved areas around the Los Angeles Maritime Museum experienced extensive cracking and settlement. In the northwest comer of the port facility, the America President Unes container terminal experienced lateral spreading, according to engineering staff of the Port of Los Angeles. A pile-reinforced dike at this location was displaced outward into the harbor, causing up to 1 foot of settlement in the backfill. The damage to the berth was extensive enough to require repair before it could be put back into service on Friday, January 21, 1994. These repairs were largely completed before the damage could be assessed by staff of the Earthquake Engineering Research Center, so the extent of the damage is unknown. Landslides The Northridge Earthquake caused scattered minor rockfalls and landslides throughout Los Angeles and Ventura Counties. Major landslides occurred in the Santa Monica and San Gabriel Mountains closing roads and destroying homes, as described below. In addition, shattered ridges were observed in the Santa Susana Mountains, north of the epicentral region. The most damaging landslides occurred in the coastal bluffs of the Pacific Palisades in Santa Monica. Here, the northbound lanes of the Pacific Coast Highway remained closed between Temescal Canyon Road and Chautauqua Boulevard for at least 4 days following the earthquake. Four large landslides were observed in this area, along with several smaller slides. These failures occurred in Quaternary and Pleistocene age deposits of weakly cemented sand (Jennings and Strand, 1969). The slopes where the failures occurred were 120 to 200 feet 4-7 in height, and moderately steep (between 45 and 60 degrees). The failure masses appeared to be only a few yards thick, subparallel to the slope, and had widths on the order of 300 feet. The slide debris was predominantly loose sand. The most damaging of these landslides occurred just north of Chautauqua Boulevard on the Pacific Coast Highway. This slide carried a portion of a house down the slope, and on adjacent properties, shallow concrete piers and H-piles were observed to be hanging in mid-air near the crest of the slope. Three homes at the crest of the bluff were condemned. Some evidence of topographic amplification of shaking was also observed in this residential development, as the most severe damage to homes tended to be at sites on the southeast comer of the bluff near the crest. Santa Susana Canyon Road is a two-lane roadway that parallels Highway 118 and connects the San Fernando Valley with Simi Valley. One section of the road dosed due to slides is approximately 5 miles from the epicenter and is built with cut slopes into cemented sand and weak sandstone canyon walls that form slopes of 2H:1V and greater. Slope failures varied from 25 feet to more than 100 feet in height. Slope failures and landslides occurred both downslope and upslope of the road. Debris from upslope failures generally was large enough to block the near lane -of the road. Blocks as large as 5 feet in diameter were noted and at least one of the slides appeared to be a failure along intersecting joint planes. Downslope failures created extensional cracks 10 to 30 feet away from the edge of the slope and parallel with the road. One larger slide caused vertical subsidence of 5 inches in the roadway and an additiona112-18 inches along the shoulder. Several earthquake-induced landslides were observed in the Angeles National Forest of the San Gabriel Mountains. Two major slides occurred at Dillon Divide, along the Little Tujunga Road that links Highways 210 and 14. No casualties were reported, but the volume of debris and the large size of the fallen rocks kept the road closed for four days. Once reopened, this road served as a main alternative access to Santa Clarita. At least 10 other slides were observed along the Little Tujunga Road. Also along the Little Tujunga Road, major pavement cracks were observed at Bear Canyon and Sand Canyon. These fresh cracks were up to 1 in wide, continuous, and hemispherically shaped, indicating deformation of the underlying fill. Retaining structures, consisting mainly of reinforced concrete crib walls, were inspected and no damage was observed. Several reinforced concrete crib walls, built by the Angeles National Forest as debris basins along Schoolhouse Canyon and the West Fork canyon, were reported to have suffered no damage by local authorities. Road crews also reported a major rockslide on Placenta Canyon Road. North of Simi Valley in the Big Mountains (near the north end of Tapo Canyon Road), landsliding occurred in an embankment adjacent to a quarry debris basin which generally contains a small amount of water. The embankment is composed of clean sand with some gravel and concrete debris. Only the base of the embankment was saturated (by water from the debris basin), and it appears that liquefaction of saturated soils near the toe may have contributed to the failure. The unsaturated soils at the top of the slope fractured into discrete blocks as they slid downslope (Figure 4.12). Workers at the quarry reported that slope movements were initiated by the main shock, but that additional movements occurred 4-8 with each of the large aftershocks. Several shallow rockslides were observed along the hills on the north side of Highway 126 between Fillmore and Interstate 5. An example of a ravelling failure in this general area is presented in Figure 4.11. Oeser inspection of a particular rockslide north of the city of Piru indicated the failure occurred in weathered sandstone. Earthquake-induced landsliding was also observed in Universal City, where a 24-foot high landslide was observed on Cahuenga Blvd. A total of nine earth or rockfill embankment dams were inspected by the EERC team after the 1994 Northridge Earthquake: Castaic, Encino Lake, Hansen, Lower Sa.n Fernando, Santa Felicia, Sepulveda, Upper Van Norman, an asphalt lined storage reservoir at the DWP water treatment facility, and a small earth embankment impounding an influent basin/reservoir. The California Division of Safety of Dams (DSOD) has undertaken an extensive inspection of all major dams within the strongly shaken region. A number of dams suffered relatively minor cracking. The Upper Van Norman Lake dam experienced minor cracking along the crest of the embankment. About twenty feet down the west side of the downstream face, three to four inches of settlement was observed around what appeared to be concrete •mini-piles•, about two to three inches in diame•er. The eastern portion of the downstream face experienced moderate cracking with Jess cracking toward the western end of the embankment. The asphalt lining of the DWP storage reservoir cracked at several locations, with one crack extending below the water level. There were also some broken pipes at the cres• of the reservoir, but it is unknown whether the earthquake caused these breaks. The small embankment at the southern (downstream) end of the influent basin at the DWP water treatment facility was breached and washed out following the earthquake. The embankment, which was on the order of 15 feet high and impounded a small pond, ~ reported to have been overtopped by DWP personnel who were at the site during the earthquake. Based on the absence of water marks on either side of the breached section and no evidence of flows higher than 2 feet in the channel below the dam, it is unlikely that the embankment was significantly overtopped. It appears likely that liquefaction of the soils below the embankment and consequent ground movements caused cracking of the embankment, which then failed due to piping and erosion, releasing the small amoun• of water in the pond relatively slowly. Evidence of liquefaction (sand boils) was observed along the west side of the basin, and evidence of slope movement was apparent immediately adjacent to the basin. The slope movement caused buckling of the shotcrete lining of the basin in two areas. In addition, the basin is located immediately downslope from the Jensen Filtration Plant, which was the site of large scale lateral spreading. There was no observed damage to Castaic Dam, but some fresh cracks were observed in the asphalt parking area on Lake Hughes Road (along the right abutment). A small slide, 4-9 50 to 60 feet wide, occurred in the area north of the right abutment and upstream of the dam, towards the reservoir. Many small surficial slides along the edge of the reservoir were also observed, and the lift gate mechanism of the outlet works was damaged. Although there was no observable damage to the San Felicia embankment, there was evidence of minor renewed slide activity in areas of previous instability along the approach road above the left abutment. Several mechanically stabilized walls were inspected along U.S. Highway 101, Interstate Highway 110, and California State Highway 2. Of these structures, signs of earthquake­ induced movements were only found along Highway 101 at Universal City. At this location, there are three large, approximately 40 feet tall walls, only one of which was damaged during the earthquake. The damaged crib wall is part of the onramp access to Highway 101 from Coral Drive. A crack parallel to the wall facing, approximately 120 feet long with 1 inch of vertical and horizontal displacement, was observed 6 feet behind the face of the wall. A second crack, also parallel to the facing, was located approximately 25 feet from the wall. The two additional cn"b walls at this location showed no signs of distress. Retaining structures were also inspected in the Angeles National Forest. These walls were primarily reinforced concrete cn"b walls, and although several landslides were observed in the region, no damage to the walls was observed. Several other earth and/or rockfill dams experienced minor cracking and distress, and several suffered damages to their abutments and/or reservoir slopes. Details can best be obtained at this early juncture through DSOD. Similarly, the Pacoima Dam (a concrete structure) suffered damages very similar to those it suffered in 1971. It is interesting to note that peak accelerations of approximated 2g were recorded at both the crest and an abutment station at Pacoima Dam. Overall, no major dams or embankments suffered significant damages posing any threat of failure, and the performance of dams was generally good. Solid Waste Landfills The 1994 Northridge earthquake provided important observational data on the response of landfills to strong levels of earthquake shaking. A large number of landfills in the Los Angeles area were located close to the epicenter and experienced strong levels of shaking, and nine of these were inspected after the event. Although no landfills demonstrated any signs of a major instability, several experienced minor levels of damage (cracking). The Simi Valley and Puente Hills landfills experienced minor cracking as a result of the earthquake shaking. At the Sunshine Canyon landfill, longitudinal cracks were observed along the crest of the waste fill. Similar minor cracking occurred at several locations on the faces of slopes of the Operating Industries Inc. (OII) Landfill, mainly at or near to berm roads. The cracks, however, did not extend through the soil cover system at these landfills, and were minor in extent, being generally on the order of 1 or 2 inches or less at their widest point and showing little or no shear offset. The cracking appeared to represent simple brittle cracking of the stiffer compacted cover soil veneers overlying the more ductile waste fill, and did not represent any threat of incipient instability. At the Lopez Canyon 4-10 landfill (shown on Figure 4.13), minor cracking was observed at the interface between the waste fill and the natural canyon slopes. Preliminary studies indicate that the relatively slight damage observed at these landfills poses no significant risk and can be easily repaired. At the Chiquito Canyon landfill, a minor amount of damage from the earthquake '\Was reported by the owner. The landfill was accepting waste at the time of the earthquake. Longitudinal cracks were observed at the crest of the landfill along the interface between the landfill liner and the waste fill. The slopes in this area were graded at approximately 2H:1V. The cracks were several inches wide with a vertical offset of several inches causing, in one area of the landfill, a small tear in the HDPE liner. The report on this landfill is preliminary and the ·site survey information will have to be closely examined before any conclusions can be drawn about its performance. In general, the performance of the major landfills, several of which appear to have been subjected to peak bedrock (input) accelerations of 0.2g to O.Sg, was very good. Acknowledgements This report represents the efforts of a large number of students, staff, and faculty at U .C. Berkeley. Numerous geotechnical professionals throughout the affected region generously provided field reconnaissance data as well as background data, and this is greatly appreciated. A number of government agencies and public utilities contnbuted to the information presented, including the California State Department of Transportation (Caltrans), the California Division of Mines and Geology Strong Motion Instrumentation Program (CDMG/CSMIP), the U.S. Geological Survey (USGS), the Los Angeles DWP, the Metropolitan Water District, and The Gas Company. We would also like to thank Geoff Martin of the University of Southern California, Phil Gillibrand of the P.W. Gilhbrand Company, Ed Kavazanjian of Geosyntec Consultants, Dr. Les Harder of the California Department of Water Resources, Dean Wise of Browning-Ferris Industries, Dean Affeldt of the PRA Group, Doug Corcorgn of Waste Management, Mike Courtemarc of Metropolitan Water District, and Anthony Shakal of CDMG/CSMIP. Tom Rockwell, Diane Murbach, Kevin Colson, and Kim Thorup of San Diego State University, Elden Goth of U.C. Irvine, Scott Lindvall of Lindvall Richter Benuski Associates, and Ken Cruikshank of Purdue University assisted in field studies of Potrero Canyon. The Newhall Land and Farming Company graciously provided access to their property. These initial investigative efforts have been supported, in large part, by the U.S. National Science Foundation, and this support is greatly appreciated. Finally, sincere appreciation is expressed to Alisa Stewart, for her tireless efforts in rapidly preparing figures for this document. References Jennings, C.W., and Strand, R.G. (1969). Geologic Map of California, Los Angeles Sheet. California Division of Mines and Geology. 4-11 ··.:.::.:,:::. ·.: :: ·::·~ ,0 Afc~0 Soil Liquefaction Oc,e-AN lateral Spreading or Ground Distress landslide Mountains Santa Monica Bay Redondo DaaAh.W. Fig. 4.1: Map of Affected Region Showing Damaged Sites 4-12 .Angeles ~, National Forest ~ ov,~~ Fig. 4.2: Extension feature at soil to bedrock contact in Potrero Canyon. Note 1-foot long ruler at right side of photograph. Fig. 4.3: Evidence of localized compression in Potrero Canyon. Note that originally straight pipe was pushed up and laterally; shortening across its length was 5 inches. 4-13 Fig. 4.4: Liquefaction of sand deposit with appreciable fines content in Potrero Canyon Fig. 4.5: Lateral spreading fissure and sand boil at dry lakebed behind Hansen Darn 4-14 Fig. 4.6: Roadway damaged by liquefaction and lateral spreading failure at dry lakebed behind Hansen Dam 4-15 Fig. 4.7: Damage from lateral spreading, Northridge Hospital Fig. 4.8: Buckling of curb from lateral spreading, Northridge Hospital 4-16 Fig. 4.9: Localized compression feature and pavement damage in Northridge near Highway 118 Fig. 4.10: Evidence of liquefaction and lateral spreading, northeast Northridge 4-17 Fig. 4.11: Bluff failure and ravelling on Highway 126 near Piru Liquefied debris Not to scale Fig. 4.12: Schematic of Slope Failure Geometry at Quarry Embankment 4-18 Fig. 4.13: View of Lopez Canyon solid waste landfill 4-19 CHAPTERS TRANSPORTATION STRUCTURES The metropolitan Los Angeles area is highly dependent on its transportation systems. Most of the 600 mile freeway system survived the Northridge earthquake with minimal or easily repairable damage. However, the extensive damage or collapse of approximately ten freeway structures caused widespread disruption after the earthquake. The structures retrofitted by Caltrans since the 1989 Lorna Prieta earthquake performed very well in most cases. Structures designed to current standards appear to have performed well, indicating that if the damaged structures had been designed to current standards many of the observed failures would not have occurred. This preliminary investigation includes reconnaissance of the freeway structures by the EERC tea111, review of available structural drawings, and tentative conclusions concerning the seismic performance of the structures. Unless stated otherwise, all the bridges described in this chapter are constructed of reinforced and prestressed concrete with multi-cell box girders (cast in place). Interstate 5/Route 14 Interchange An overall view of the Interstate 5/Route 14 interchange in Fig. 5.1 shows that the most significant damage was in the North Connector Overcrossing (west 14 to north 5) and the South Overhead (west 14 to south 5). The connectors are box girders supported by single column bents. While under construction, several structures in the interchange collapsed or were damaged in the 1971 San Fernando earthquake. The North Connector was repaired shortly after that earthquake. North Connector (Bridge No. 53-1964F) The eastern end frame of the North Connector collapsed, as shown in Fig. 5.2. The frame consists of a simply supported span between abutment 1 and bent 2 and continuous spans over bents 2 to 4. The simple span fell off the seat abutment, but the transverse shear keys remained intact. A shear failure in the bent 2 column appears to have initiated the collapse (Fig. 5.3). The 4 ft by 8ft column has 42 #18 bars longitudinally and #4 ties at 12 in. with 3 #4 crossties. The column is roughly one-half as tall as the adjacent column at bent 3 and, therefore, probably attracted larger shear forces. After the shear failure of bent 2 and resulting loss of gravity load capacity, the box girder formed a hinge at bent 3 and subsequently tore out. One of the restrainer units at the hinge near bent 2 pulled out the diaphragm, and in the other unit the restrainers pulled out of the bearing plate. The next hinge, near bent 4, was barely providing support on about 2 in. of the 14 in. seat. South Overhead (Bridge 53-1960F) As shown in Fig. 5.4, the southern end frame of the South Overhead collapsed. The frame consists of a seat abutment and continuous spans over bents 2 and 3 with a hinge near bent 4. The shear crack in the bent 3 column (Fig. 5.6) indicates motion towards the abutment. The column at bent 2 most likely failed first because it is only about one-third the height of bent 3 and, therefore, would have attracted a larger shear force than bent 3. The loss of bent 2 could have caused the box girder to hinge at that location as evidenced by flexural cracks at the bottom of the girder (Fig. 5.5), and pulled the girder off the abutment. The increased gravity load at bent 3 then appears to 5-1 have caused the cap beam failure, with the subsequent punch through of the column, and pulled the box girder off the hinge near bent 4. The splitting of the bent cap into a wedge shape along bent 3 could be due to flexural hinging of the deck as it collapsed on either side. The top reinforcement from the bent cap lay unbent across the deck near the column base. Interstate 5/Route 118 Interchange, Southwest Connector (Bridge No. 53-2329) Fig. 5.7 shows bent 2 in the end frame just before the connector crosses Sharp Ave. The column experienced large longitudinal forces as evidenced by the shear cracks and soil displacement on the east side of the column base. The reinforcement of the 8 ft octagonal column consists of 64 #11 longitudinal bars and #4 spirals at 3.5 in. pitch. The incipient shear failure of the column may be due to larger forces at the stiff end frame or a higher point of maximum moment in the CIDH (cast in place drilled hole) foundation than assumed for the design in 1972. The hinges, with 2 ft seat width, showed evidence of pounding, and abutment 1 was damaged. Interstate 5/210 Interchange, Southwest Connector (Bridge No. 53-1989F) This connector, which carries traffic from east 210 to south 5, is part of an interchange that was heavily damaged in the 1971 San Fernando earthquake. The connector has seven single column bents, seat abutments, and no intermediate hinges. In the Northridge earthquake, both abutments suffered extensive damage. At abutment 1 (northeast end) the box girder pulled out at least 3 in. and there was evidence of pounding longitudinally and transversely. Abutment 9 (southwest end) had extensive pounding damage and it spalled in a pattern consistent with twisting of the box girder. Bent 2, which is very short, had considerable spalling at the top and bottom of the column. Flexural spalling and cracking was observed at bents 3, 4, and 5. The column for bent 6 passes through an opening in the box girder of the San Fernando Undercrossing (Bridge No. 53-1730). The nominal 6 in. gap between the column and undercrossing deck pounded during the earthquake, most likely due to displacement of the deck (see below). Interstate 5, San Fernando Road Undercrossings Interstate 5 crosses over San Fernando Road at the interchange with Interstate 210 (Bridge No. 53-1730) and again further north near the interchange with Route 14. At both locations the undercrossing is skewed, and there is evidence of pounding and pullout at the abutments and damage to the wingwalls. The intermediate hinges also showed slight pounding damage. Several columns in the multi-column bents had minor spalling near the soffit Interstate 405/Interstate 10 Interchange The connectors and overpasses at the Interstate 405/10 interchange were retrofitted with full­ length steel jackets and hinge restrainers, and the foundations were strengthened. Inspection of the structure showed relative movement and pounding at most of the hinges. The connector from westbound Interstate 10 to southbound Interstate 405 experienced shear cracking of the girder seat and vertical and horizontal offset of the roadway at the hinge atop aT -shaped column, as shown in Fig. 5.8. There was no visible damage to the retrofitted northbound Interstate 405 to westbound 5-2 Interstate 10 connector. Strong motion instruments at that connector showed a peak vertical acceleration of 1.83 g as the box girder separated and pounded on the abutment. Route 118, Bull Creek Canyon Channel Bridge (Bridge No. 53-2206) The two bridges were designed in 1973. The westbound structure is supported by a four­ column bent (bent 2) and a five~olumn bent (bent 3). The eastbound structure has the same geometry, but both bents have five columns. The abutments are at different skew angles. The transverse reinforcement in the columns is #5 smooth spirals at 4 in. pitch for one column diameter at the top and bottom, and 12 in. pitch in the middle. The two southernmost columns of bent 2 showed plastic hinging in and below the confined length, with fractured spirals and buckled longitudinal bars (see Fig. 5.9). The soffit adjacent to the hinged bent 2 columns were spalled with large cracks parallel to the bent. The retaining wall for the channel is located directly against the base of all the columns in bent 3. As a result of the restraint by the channel wall, all coluoms along bent 3 failed in shear just above the wall and the confined zone, as shown in Fig. 5.10. Both abutments appeared undamaged from below. However, the approach slab on the southeast side had been pulled south about 13 in. This may indicate that the abutment or backfill was flexible enough to allow the structure to displace transversely, increasing the forces on the columns. Route 118, Mission-Gothic Undercrossing (Bridge No. 2205) The Mission-Gothic Undercrossing consists of two parallel structures designed in 1973. The abutments have a 90 degree difference in skew because of the intersection of the two streets below, as shown in the schematic plan in Fig. 5.11. The westbound structure is 506 ft long with two bents, and the eastbound structure if 566 ft long with three bents. The 6 ft octagonal columns are flared at the top. During the main event and aftershocks, the eastbound structure came off the east abutment and collapsed. The westbound structure partially collapsed but remained on the abutments (Fig. 5.12). The earthquake displaced both columns in bent 3-left of the westbound structure transversely, with a plastic hinge forming below the flares (Fig. 5.13). The columns in bents 4-right and 3-right of the eastbound structure also had plastic hinges below the flares. There was additional damage to shearing deformation in the hinge. The two columns forming bent 2-left of the westbound structure were displaced in the longitudinal direction, producing hinging about the weak axis of the flare near the soffit. At the eastern abutment supporting the westbound structure, the recessed shear keys had completely spalled, while the raised shear keys were not damaged. The pattern of abutment damage and column deformation seems to indicate that the eastbound structure rotated about the western abutment. The westbound structure also rotated about the western abutment, but to a lesser degree, probably because of its shorter length. Interstate 10, Fairfax-Washington Undercrossing (Bridge No. 53--1580) The damaged region of this undercrossing is between the west abutment and the first hinge (Fig. 5.14). The frame is supported by a seven~olumn bent Shear cracking, compressive crushing of the concrete, and symmetrical longitudinal bar buckling were evident in all the columns of the bent, as shown in Fig. 5.15. The columns at the next bent, east of the hinge, had diagonal shear cracks. As a :result of the crushing of the columns, the box girder lifted off the west abutment :rockers, and formed plastic hinges in the girders near lap splices of the girder 5-3 reinforcement The adjacent Cadillac Undercrossing, which steel jackets, suffered no visible damage. been retrofitted with full-length Interstate 10, La Cienega-Venice Undercrossing (Bridge No. 53-1609) This collapsed undercrossing consists of two and two pier walls (Fig. 3 and and bents 6 and they (,'Tacking, compressive crushing of concrete, and symmetrical longitudinal evident in most of the columns for the westbound structure, as shown in deck unseated at the hinge between bents 6 7 and apparently the for girder reinforcement. eastbound structure to a lesser extent, possibly a nearby ramp structure. damaged columns revealed transverse reinforcement of #4 spliced hoops at 12 in. damaged columns. the hoops had either fractured or the lap splices had opened. 5, Interstate 5 crosses Gavin Canyon on two skewed five-span bridges (Fig. 5.19) constructed 1967. central of each structure is prestressed and beyond bents to provide for the adjacent an 8 seat at hinge. cracking at the bases of the columns supporting the center spans suggested east-west displacement or rotation in the plan of the superstructure. It appears that the displacements were sufficiently large to cause loss of bearing support at the hinge seats. The unseating apparently caused flexural failure of the box girders. Cable restrainers had across the seats in a retrofit Route 405, Jefferson Blvd. Undercrossing (Bridge No. 53-1255) The undercrossing was and in During later construction of the Route 90 interchange, columns were the outriggers supponing Interstate 405. During the earthquake, shear cracks in both directions extending into the lower column were evident on all the exterior outrigger joints of the double-deck columns (Fig. 5.20). Connectors with single column bents had been with full--length steel jackets. was no visible in retrofitted connectors. Interstate 5, Santa Clara River Bridge (Bridge No. 53-0687) required to open a 5 in.-diameter pipe rail at the west side of the northbound structure. The observed damage was consistent with a clockwise rotation of superstructure frames. Route 101, Los Virgenes Overcrossing (Bridge No. 53-1442) Los Virgenes crosses over Route 101 on a four-span bridge constructed in 1974. The bridge consists of a concrete deck over steel girders supported on three multi-<;olumn concrete bents and concrete abutments. There are hinges in the superstucture at the first and last bents and along the centerline of the deck. Pounding damage was observed at the hinges. Abutment flll settlements of approximately 4 in. were observed. Caltrans employees excavated soil around piles supporting the south abutment. Damage to a pile at the east end of the abutment was observed (Fig. 5.23); no damage was observed at the west excavation. The potential for damage to piles in other bridges with relatively minor superstructure and abutment damage may be worth further investigation. Interstate 5/Route 126 Separation (Bridge No. 53-1626) Route 126 west diverges from Interstate 5 on two four-span bridges constructed in 1964. Each structure is a box girder supported on three-<;olumn bents and the abutments. The columns showed evidence of rocking and each had spalled at the southeast comer near the box girder soffit Concrete had spalled and vertical steel was exposed and deformed at the abutment connection to the box girder at the east abutments. No structural damage was present at the west abutments. Ground cracks were observed at the eastern median strip and on the eastern abutment fill slopes. These cracks increased in size to the east of the bridge, approaching 2 inches in width, 2 to 3 inches in vertical offset, and tens to hundreds of feet in length (Fig. 5.24). The abutment fill at the west approach slab of the northern roadway settled about 2 in. Other Damage Sound walls constructed on concrete barriers along the northern edges of Route 101, just east of Interstate 405, were observed in various conditions. Some walls, shown in Fig. 5.25, have remained vertical, others have rotated, and another has collapsed to the south. The collapsed wall has #5 longitudinal reinforcement placed in every fourth cell (approximately 32 in. on center) of a precast masonry panel. Lap splices of approximately 28 in. in length begin at the base of the wall. Concrete had been placed in the cells containing longitudinal reinforcement with one exception. At every location the wall reinforcement had failed near the connection to the concrete barrier rail. Approximately one-half of the bars failed by fracture. Signs of necking were not observed, and the fractures occurred as far as 10 in. below the wall/barrier interface. The remaining bars had lap splice failures. The precast masonry was disrupted at the end of the wall where a lap splice failure was observed. As of 9:51 a.m. on 19 January 1994, Caltrans reported damage to approximately thirty additional structures. Most of the reports consisted of approach slab settlement, abutment damage, bearing damage, or minor column spalling. A later report indicated damage to connectors at the Route 134/2 interchange which had been retrofitted with steel jackets. The City of Los Angeles reported damage to a pedestrian overcrossing at Wilbur Ave. in San Fernando, in addition. to settlement at approach spans and shear key damage in several city bridges. Railroad companies reported inconsequential damage and nearly full operation of their bridges. 5-5 than nee:ae~:J., however. to ,,.,., . .,."lr"T...,, unseating failures. The post-earthquake Caltrans. Immediately lPn~rti'Y!Plr'lt of 5-7 Ul Jo Fig. 5.2: Overview of Interstate 5/Route 14, North Connector Fig. 5.3: Bent 2 of Interstate 5/Route 14, North Connector Failure Fig. 5.4: Overview of Interstate 5/Route 14, South Overhead Failure Fig. 5.5: Bent 2 of Interstate 5/Route 14, South Overhead Fig. 5.6: Bent 3 of Interstate 5/Route 14, South Overhead Fig. 5.8: Seat Failure at Connector from West Interstate 10 to South Interstate 405 5-9 Fig. 5.7: Bent 2 of Interstate 5/Route 118 Southwest Connector Fig. 5.9: Column Failure at Bent 2, Bull Creek Canyon Channel Bridge Ul I -0 Fig. 5.10: Column Failure at Bent 3, Bull Creek Canyon Channel Fig. 5.11: Schematic Plan of Mission-Gothic Undercrossing Bridge Fig. 5.12: West View of Mission-Gothic Undercrossing Fig. 5.13: South Column of Bent 3-Left, Mission-Gothic Undercrossing 'f ..... ..... Fig. 5.14: Overview of Fairfax Undercrossing Collapse Fig. 5.16: Overview of La Cienega-Venice Undercrossing Collapse Fig. 5.15: Column Failure at Fairfax Overcrossing Fig. 5.17: Column Failure at Bent 7, La Cienega-Venice Undercrossing Ul I ...... N Fig. 5.18: Column Failure at Bent 6, La Cienega-Venice U ndercrossing Fig. 5.20: Joint Failure at Jefferson Blvd. Undercrossing Fig. 5.19: Collapsed Spans at Gavin Canyon Undercrossing Fig. 5.21: Pier Wall at Southbound Span of Santa Clara River Bridge Looking South VI I .... w Fig. 5.22: Restrainer at Santa Clara River Bridge Fig. 5.23: Pile Damage at Los Virgenes Overcrossing Fig. 5.24: Ground Cracks at East Approach of Interstate 5/Route Fig. 5.25: Sound Walls at Route 101 near Interstate 405 126 Separation CHAPTER6 BUILDING STRUCTURES General The January 17, 1994 Northridge Earthquake affected a very large and densely populated urban and suburban area with a wide range of structural types. While the epicenter was located in the city of Northridge in the San Fernando Valley, high horizontal accelerations were recorded at sites as far as 36 km from the epicenter in downtown Los Angeles. Several free field instruments at a distance between 10 and 15 km in the north/north-east direction from the epicenter registered very high horizontal and high vertical accelerations: 0.91g horizontal acceleration at the Sylmar Country Hospital and 0.59g vertical acceleration at the Nordhoff Avenue Fire Station. The large number of buildings and structural types that were affected by the earthquake make a thorough evaluation impossible at this early stage. This preliminary report contains information on the following types of structures for which several instances of significant structural or non-structural damage, partial or total collapse were observed: reinforced concrete buildings, parking structures, hospitals, unreinforced masonry buildings, wood residential structures and base isolated structures. Parking structures had a large incidence of partial and total collapse cases among modem engineered structures and are, therefore, treated separately and at some length in this report. Hospitals are also treated separately, because of their importance and the high incidence of non-structural damage that was observed. No reports of significant damage to steel buildings were obtained at this early stage. Reinforced Concrete Buildings Reinforced concrete buildings suffered significant structural damage. Two buildings suffered partial collapse, while another suffered such serious structural damage that it had to be immediately demolished. The information in this report is based on structures located in Shennan Oaks (approx. 10 km from the epicenter), Culver City (approx. 25 km from the epicenter) and Santa Monica (approx. 25 km from the epicenter). All types of structural systems from ductile frames, shear walls, coupled shear walls and dual systems to non-ductile reinforced concrete frames suffered structural damage. The observed types of failure in reinforced concrete buildings include shear cracking and compression spalling in poorly reinforced beam-column joints in the moment-resisting frame in Fig. 6.1, shear and bond-splitting failure in the column ends of the third story of the 7-story ductile moment-resisting frame in Fig. 6.2, shear failure of poorly detailed beam and column ends and beam-column joints in the five story non-ductile reinforced concrete frame in Fig. 6.3 which suffered partial collapse, short column shear from the second to the fifth floor in the six story frame in Fig. 6.4, splice failure at the comer shear wall of the dual wall-frame system in Fig. 6.5 and shear failure of the coupling beams in the 15-story coupled wall system in Fig. 6.6. 6-1 Fig. 6.1 is representative of the type of problem that was observed in several cast in place reinforced concrete parking structures which are discussed in detail in the following section. The 7-story moment resisting frame in Fig. 6.2 is located in Van Nuys approx. 7 krn east of the epicenter and represents the closest instrumented building. It was built in the mid sixties and suffered non-structural damage totaling $400,000 during the 1971 San Fernando earthquake under a peak horizontal acceleration of 0.27g. [1]. During the Northridge earthquake it experienced a peak horizontal acceleration of 0.47 g at the base and 0.59g at the roof. The strong ground motion lasted for about 15 sec and included a significant peak venical acceleration of 0.30g which appears to precede the strong part of the horizontal ground motion by about 5 sec. The significant vertical component of the ground motion that preceded the horizontal component is evident in several records in the vicinity of the epicenter. The building suffered serious structural damage in all columns of the third floor with signs of shear-bond splitting type of failure. Fig. 6.3 shows a detail of the non-ductile reinforced concrete frame of the 5-story Kaiser Per