Subsurface geotechnical investigations near sites of ground deformation caused by the January 17, 1994, Northridge, California, earthquake

This report contains geotechnical and geologic data from 62 CPT soundings and 29 hollow-stem auger borings performed at four sites of permanent ground deformation caused by the 1994 Northridge, California, earthquake. Three of the sites were in the epicentral region within the San Fernando Valley, the fourth site was in Potrero Canyon. All four sites experienced high levels of ground shaking and ground cracking and were in areas where underground utilities, gas transmission lines, and foundations were damaged. The subsurface investigations found materials with a low resistance to liquefaction at three of the sites Balboa Boulevard, Wynne Avenue, and Potrero Canyon and a weak soft clay at the fourth site Maiden Street. INTRODUCTION The January 17, 1994, Ms=6.8 Northridge, California, earthquake created an estimated property loss of between $20 and 30 billion (Holzer and others, 1996), and was responsible for 57 fatalities and more than 9,000 people injured (Gordon, 1994). The earthquake occurred on a blind reverse fault beneath the San Fernando Valley. No primary surface faulting was associated with the earthquake (Gordon, 1994). The widespread damage to buildings, roads and life lines is attributed to the location of the epicenter within an urban region and the unusually high ground motions for a California earthquake of this size. Peak accelerations ranged from 1.8 g, at Tarzana (7 km from the epicenter) to 0.9 g at Sylmar (15 km from the epicenter) to 0.5 g at several other sites within 40 km of epicenter (Peterson, 1994). Ground cracks and ground failures in the alluvium underlying the San Fernando Valley were extensive during the earthquake but the mechanism of failure at some sites was unclear. This report contains subsurface data that we collected to help resolve the questions as to the mechanisms of failure. A preliminary report was published by Holzer and others (1996). Holzer and others (in press) present a comprehensive analysis of these data. Investigations were conducted at four ground failure sites during 1995 and 1996. The investigation consisted of cone and standard penetration tests, in-situ vane tests, and Shelby tube sampling. Laboratory tests on samples included: grain size, water content, bulk density, Atterberg limits, and laboratory vane tests. These investigations also were augmented with other USGS investigations described in Holzer and others (in press). METHODS Cone Penetration Test (CPT) Cone penetration tests were made at each of the four sites to define stratification and measure penetration resistance for use in estimating liquefaction resistance. The cone used is a 10-ton subtraction cone with a single element strain gauge in a 3.6 cm diameter housing. The 60-degree conical tip has an end area of 10 cm2 . The sleeve behind the tip has an area of 150 cm2 . The cone was advanced into the ground at a rate of 2-cm per second. The procedures and equipment follow 1 the requirements described in the American Society of Testing and Materials (ASTM)guidelines D-3441-79 (ASTM, 1983). Standard Penetration Test (SPT) Following CPT soundings, SPT borings were made about 1.5 m away to measure dynamic penetration resistance for liquefaction analyses and to obtain samples for soil index tests. The SPT procedures follow the ASTM guidelines in D1586-67 (ASTM, 1983). These procedures were modified (Youd and Bennett, 1983) for use with hollow-stem augers (25-cm outside and 10-cm inside diameter). A Mobile2 split barrel "ADO standard penetration sampler" (5 cm outside and 3.5 cm inside diameter) was used with liners. The sampler was driven into the soil with a Mobile "In-hole sampling hammer" (64 kg) falling 76 cm. The hammer is raised and dropped using a Mobile "Safe-T-Driver" hoist. After the hydraulic system has lifted the hammer the required distance the winch is reversed and cable is thrown off the hoist allowing the hammer to fall freely. The penetration resistance (N) is equal to the number of hammer blows to advance the sampler 30 cm after an initial seating of 15 cm. The measured overall efficiency of this system is 68 percent (Douglas and Strutynsky, 1984). Samples were described in the field (including color), water content samples were sealed in containers, and carbon samples for dating were also collected. Undisturbed samples were taken with thin-walled Shelby tubes (7.6 x 91 cm and 8.9 x 76 cm). Tubes were slowly pushed into the soil past the tip of the auger, the auger was then advanced the sampling length, following which the tube was then rotated to shear off the sample from the undisturbed soil. In-situ field vane tests were made through the hollow-stem auger to a maximum depth of about 6.4 m. The hand held vane instrument is a Pilcon model DR 799. The vane was rotated at 90 degrees per minute. This is a faster than recommended rate, but because the device is hand held and operated without a gearing system it was a rate that could be reproduced with confidence. Justification for the high rate is described in Holzer and others (in press). Laboratory Methods Index tests conducted in the laboratory include; grain size (D422-63, ASTM, 1983), liquid limit (D423-66, ASTM, 1983), plastic limit (D424-59, ASTM, 1983), and water content (D2216, ASTM, 1983). Samples were classified using the Unified Soil Classification (D2488-69, ASTM, 1983) as modified by Howard (1984). Shelby tube samples were cut into 15 cm lengths for bulk density and vane tests. The length of each subsample was measured and the tube and sample weighed. A pocket penetrometer was then used to measure compressive strength at the end of the sample. Next, shear strength was measured using a Wykeham Farrance laboratory vane shear with the sample still in the tube. The vane (12.7 x 12.7 mm) was pushed into the sample 40 mm and rotated at 90 degrees per minute. Peak strength, residual strength and remolded strength were determined, then the vane was removed and a water content sample was taken from the area of the vane test. The sample was then extruded and described. Results of the laboratory tests are 2Use of trade names is for descriptive purposes only. shown in Table 2. CPT profiles, SPT data, median grain size and fines content, and brief field descriptions are shown in the logs. Radiocarbon Analyses Radiocarbon dates were determined using the accelerator mass spectrometer (AMS) technique. The reported dates have been adjusted by carbon-13 for total isotope effect generated in both nature and during the physical and chemical laboratory procedures. The carbon-13 content was measured concurrently with that of carbon-14 and carbon-12 in the accelerator beam, allowing a precise correction. These dates are reported as RCYBP (radiocarbon years before 1950 A.D.). By international convention, the half-life of radiocarbon is taken as 5568 years and 95% of the activity of the National Bureau of Standards Oxalic Acid (original batch) used as the modern standard. The quoted errors are from the counting of the modern standard, background, and sample being analyzed. They represent a one standard deviation statistic (68% probability), based on the international convention, no corrections are made for DeVries effect, reservoir effects, or isotope fractionation in nature, unless specifically noted (Table 4). Stable carbon ratios are measured on request and are calculated relative to the PDB-1 international standard; the adjusted ages are normalized to -25 per mil carbon 13 (Bennett and Tinsley, 1995). Liquefaction Resistance Liquefaction resistance (figs. 28 and 29) was calculated using guidelines developed in the NCEER workshop on evaluation of liquefaction resistance held in Salt Lake City, Utah, January 4 and 5, 1996 (Youd and Idriss, 1997). These guidelines update the empirical work of Seed and others (1985). The NCEER guidelines were incorporated into an Excel spreadsheet by Sam Gilstrap of Brigham Young University (written communication, 1996). Liquefaction resistance is shown in Table 3. Holzer and others (in press) discuss the levels of ground motion used in the analyses (Table 3) and they also describe the overall approach used in the analysis. SITES The first phase of the investigation began in May 1995 when CPT soundings were obtained at each site. Once the stratigraphy had been determined, SPT's were conducted to obtain samples and blow counts for liquefaction analysis. The first phase concluded in August 1995. The second phase started and ended in June 1996. CPT's and SPT's were conducted in areas outside the failure zones at Wynne Avenue and Maiden Street to determine with more certainty the subsurface conditions outside the failure zones. Balboa Boulevard The Balboa Boulevard site is located in the northern end of the San Fernando Valley where a 5-km wide complex of ground cracks formed. This complex of cracks was the most damaging ground failure to occur during the earthquake (Hecker and others, 1995). Foundations of homes were damaged and buried utilities were ruptured. One very large gas main in the middle of Balboa Boulevard was ruptured and the gas subsequently caught fire, destroying several homes. Our investigation was conducted in an unnamed alley (parallel to Balboa Boulevard) about 40 m west of Balboa Boulevard. The alley is in the western section of the main zone of cracks. These cracks run generally perpendicular to the regional topographic slope of 1.6%. The Balboa profile line was about 570-m long, and included 17 CPT soundings and 13 borings. Borings were used for SPT's, Shelby tube samples, field vane tests, and water monitoring wells. The monitoring wells were used to check the stability of the ground water level. The locations of the soundings and borings are shown in figures 1, 2, and 3. Maiden Street The Maiden Street site is located in a vacant lot that was intersected by a 0.5-km-long system of cracks that broke lifelines and damaged the foundations of several houses. There the crack system trends genera