Design and Construction of a Full-scale 5-story Base

1 Department of Structural Engineering, University of California, San Diego, 9500 ... In this paper, the design and construction of the test building...
17 downloads 714 Views 2MB Size
Download PDF
Love PNG Images
Design and Construction of a Full-scale 5-story Base Isolated Building outfitted with Nonstructural Components for Earthquake Testing at the UCSD-NEES Facility Chen, M.1, Pantoli, E.1, Wang, X.1, Espino, E.2, Mintz, S.1, Conte, J.1, Hutchinson, T.1, Marin, C.3, Meacham, B.4, Restrepo, J.1, Walsh, K.2, Englekirk, R.5, Faghihi, M.5, and Hoehler, M.6 1 Department of Structural Engineering, University of California, San Diego, 9500 Gilman Drive, Mail Code 0085, La Jolla, CA 92093; email: {mcchen,epantoli,xiw002,sjmintz,jpconte,tara,jrestrepo}@ucsd.edu. 2 Department of Civil, Construction, and Environmental Engineering, San Diego State University, 5500 Campanile Dr., San Diego, CA 92182; email: [email protected], [email protected] 3 Department of Civil Engineering, Howard University, 2300 Sixth Street, NW Washington, DC 20059; email: [email protected] 4 Department of Fire Protection Engineering, Worcester Polytechnic Institute, 110 Institute Road, Worcester, MA 01609; email: [email protected] 5 Englekirk Structural Engineers, 2116 Arlington Avenue, Los Angeles, CA 90018; email: {robert.englekirk, mahmoud.faghihi}@englekirk.com 6 Hilti Corporation, Corporate Research and Technology, Feldkircherstrasse 100, Schaan, 9494, Liechtenstein; email: [email protected] Abstract This project involves earthquake and post-earthquake fire testing of a fullscale five-story building furnished with nonstructural components and systems (NCSs). A broad array of NCSs are incorporated in the test building, including a functioning passenger elevator, stairs, exterior walls, interior partition walls, piping, HVAC, ceiling, sprinklers, building contents, as well as passive and active fire systems. The NEES-UCSD outdoor shake table facility is utilized to support this fullscale testing program. In this paper, the design and construction of the test building and its NCSs are summarized. Testing is projected for March 2012, and pending this progression, preliminary data may be presented at the conference. Data obtained from this program will be used to evaluate and refine current code assumptions and computer models regarding nonstructural components and systems, and to find ways to minimize undesirable damage to NCSs during an earthquake. Introduction Seismic damage induced on nonstructural components and systems (NCSs) continues to plague society, leading to undesirable consequences including excessive economic losses, downtime, and harm to humans. For example, the 2010 Baja California earthquake resulted in four fatalities and 100 injuries mostly from falling nonstructural components in buildings (Hutchinson et al., 2010). Observed NCS damage included collapsed parapets, damage to glazing, cracked walls, fallen soffits, failure of unreinforced masonry walls, and partial collapse of roofs. Damage extended to schools in the U.S., and included damage to doors, fallen bookshelves,

light fixtures and ceilings, some remaining out of service for months with costly repairs. Likewise, the 2010 Chile earthquake imposed significant nonstructural damage to all types of buildings (Moehle et al., 2010). The closure of airports due to nonstructural damage cost the airline $25 million in downtime, while 62% of the hospitals in the affected regions of Chile suffered from nonstructural damage, causing most of the hospitals’ economic losses, evacuations, and closures. Regardless of the continued evidence of their vulnerability, to date, only a handful of building-NCS system level experiments have been conducted. This has been largely due to cost and scale limitations (limited funds to support such extensive full-scale tests, and limited availability of facilities to support adequately sized test structures). With NCSs encompassing around 80% of the total investment in building construction (Miranda and Taghavi, 2003), and with the majority of earthquakeinduced direct losses in buildings attributed to NCS damage, it is important to complement our existing experimental database with full-scale integrated tests. By conducting an integrated (building-NCS) test, inertial forces and deformations generated by the earthquake motion to the building will be transferred to the attached and unattached NCSs. Amplified forces generated by the dynamic response of the building can be detrimental to the NCSs, even at levels that do not damage the structure itself. To address the shortcomings in knowledge, to-date several full-scale buildings have been tested at the E-Defense shake table in Japan and the NEES@UCSD shake table in the United States. In a few of these, NCSs were implemented, however, the inventory and scope of the NCS was secondary to other objectives of the test program. For example, a full-scale four-story steel building tested at E-Defense included lightweight concrete external cladding panels, aluminum sash, glass windows, gypsum board partition walls, a hanging ceiling system and more (Matsuoka et al., 2008). In addition, a full-scale five-story steel moment frame building tested at E-Defense was outfitted with interior walls, ceilings, piping, and concrete cladding panels. This test focused on displaying the effectiveness of using friction pendulum isolators to protect the structure and its contents during extreme ground shaking (Dao et al., 2011). Another full-scale test performed at E-Defense included the testing of a four-story base isolated RC hospital building. Nonstructural components, including a variety of furniture, medical appliances and service utilities, representing typical hospital layouts were placed in this structure. The purpose of this test was to demonstrate the effectiveness of protecting medical facilities using base isolators (Sato et al. 2011). The seven-story RC building tested at the NEES@UCSD facility incorporated a suspended pipe system and evaluated the forces transferred to the system’s anchorage during seismic motions (Hoehler et al. 2009). In conventional nonstructural test setups, the components are either tested individually or in subsystems, and are mounted directly to the shake table and subjected to anticipated floor accelerations and deformations (e.g. Mosqueda et al. 2009). Scope of this Work This landmark project involves the testing of a full-scale five-story building subjected to a range of earthquake motions. By constructing a full-scale building, a

broad variety of NCSs, which would commonly be installed to support the services and occupancies of a building, are implemented into the structure. With the full-scale and integrative nature of the project, the outcome is a unique investigation of the interaction between the structural and nonstructural components in realistic seismic conditions. The testing sequence will consist of first isolating the building, subjecting the structure to a series of serviceability-level earthquake motions, then jacking the entire building up and removing the isolators. The test program will continue by lowering the building and post-tensioning the foundation down to the table to support seismic testing in a fixed base configuration. Testing is conducted at the NEES site at the University of California, San Diego (Figure 1). This site provides the opportunity to construct tall, full-scale structures due to its outdoor nature. The outdoor shake table is the first and largest in the United States and can impose a lateral acceleration of 4.2g with no load. A fullscale testing site is essential for capturing responses that cannot be captured with a smaller scale.

(a) (b) Figure 1. (a) View of NEES site @ UCSD and (b) shake table specifications (NEES@UCSD, 2011) By achieving a better understanding of the total building systems performance during an earthquake, data obtained in this test can be used to evaluate nonlinear simulation tools and current code assumptions regarding nonstructural components and systems, as well as develop a more precise and specific set of performance objectives that cater to the clients’ exact needs. There is a strong need to evaluate and refine current design limit states for different performance objectives. By monitoring the performance of the NCSs during a series of ground motions, we will obtain the needed data to accomplish these goals. This paper will present an overview of the building design and construction, and a description of nonstructural components and systems. Building Design The building is designed assuming a high seismic zone in Southern California, namely downtown Los Angeles, with site class D soil conditions corresponding to “stiff soil.” Using global and system level performance criteria, seven MCE ground motions and three serviceability motions based on the site criteria are spectrally

matched and chosen for the design of the fixed base structure. The building is designed based on a performance target of around 2.5% maximum lateral interstory drift ratio with a maximum peak floor acceleration of around 0.7g-0.8g. Figure 2a shows the spectral acceleration graph for the seven MCE ground motions used for design of the structure, however the building will be tested for only one of the MCE motions. Figure 2b shows the acceleration time history for the Denali motion, which is preliminarily selected in the test program as the maximum motion due to its long duration of strong shaking and broad spectral characteristics.

(a) (b) Figure 2. Elastic 5% damped spectral acceleration of the seven MCE level design motions. The cast-in-place reinforced concrete test building has a plan dimension of 21.5 feet by 36 feet and is composed of two bays in the shaking direction and one in the non-shaking direction (Figure 3). The floor-floor height is 14 feet with a total height of the bare structure being 75 feet above the shake table (~88’ including the roof mounted equipment). A pair of moment resisting frames is incorporated on the longitudinal ends of the building. The moment resisting beams are designed with equivalent moment capacities, however different types of detailing are used at different levels. Floors two and three have high strength steel (MMFX) frame beams, with a yield strength of 120ksi. Floor four has an upturned hybrid SMF beam with four post-tensioned tendons and Dywidag ductile rod connectors. The fifth floor has a Dywidag Ductile Connector (DDC) frame beam, which utilizes ductile rod connectors. Details adopted on levels one through five articulated stable ductile performance in previous test programs (Warcholik and Priestley, 1998a,b; Chang et al., 2009) and have since observed use in practice. Lastly, the roof level, which expects the lowest forces, is detailed using current special moment resisting frame detailing of ACI. Strain gauges are placed on select beam ends where the highest strains are expected. An 8” thick cast-in-place slab provides the floor support at each level. The columns are detailed with a prefabricated hoop and transverse reinforcement grid. Using this technique, time is saved in assembling the column cages. Openings of 7’-8”x13’-9¼” and 6’-11”x8’-8” are provided in the floor diaphragm to accommodate full-height stairs and an elevator, respectively. Nominal (4”-10” diameter) penetrations for building services (plumbing, fire sprinklers, etc.) are also provided. A 6” thick concrete wall is constructed on either side of the

elevator opening running transversely to provide gravity support for the elevator guiderails and torsional stability during the seismic tests (Figure 3b).

(a)

(b)

Figure 3. Schematic overview of (a) building elevation and (b) plan view. The foundation is designed with a unique solution using a combination of post-tensioned rods and tendons (Figure 4). The foundation is designed to remain elastic such that its dynamic behavior does not affect the performance of the nonstructural components and systems. A rigid foundation is also important for being able to make direct comparisons between the base-isolated and fixed-base conditions.

Figure 4. Reinforcement details of the longitudinal foundation beams. Construction Construction commenced in early May 2011 with the foundation formwork. Temporary rigid steel tubes were used below the building for its support during construction in lieu of the rubber isolators. Following placement of foundation formwork, tendons, rods, reinforcement, and the temporary support blocks, one hundred cubic yards of foundation concrete were poured. A shoring system was then assembled and reinforcing steel placed at each floor at a rate of approximately two weeks per floor between concrete placements. Each floor’s concrete pour was divided into three days, the first day being the slab pour, the second including the two

elevator walls and interior columns, and the third including the four corner columns. Concrete pours were divided due to the physical characteristics of the formwork and the layout of the structural elements. Figure 5 shows images of construction of the building at various stages.

(a) (b) (c) Figure 5. Photographs of test building during construction (a) foundation beams and level 1 columns, (b) level 3 deck, and (c) level 5 deck. The general contractor used one floor of shoring, and two floors of reshores immediately below. Once concrete reached five-day strength, the reshores were removed and placed on the active slab above to begin the next level’s shoring. This process was supported with an onsite mobile crane, used for both jumping the scaffolds and concrete placement via a bucket. The scaffolding system used for the project was a floating system. The scaffolding was attached to the columns, and was lifted up with a crane as a unit as the construction progressed (Figure 6). The bare structure was completed on September 23, 2011. System identification using white noise tests conducted on the bare structure indicate the first and second modes of the bare structure were longitudinal and transverse with values of 0.57 and 0.53 seconds and 4.08% and 1.04% damping, respectively.

(a) (b) Figure 6. Photographs of the construction process: (a) floating scaffolding and (b) moving scaffolding system to next level.

Nonstructural Components and Systems Each level of the building is designated with a different occupancy and as such is outfitted with different types of nonstructural components and systems. The first floor is considered the utility floor and also has several systems that represent typical “East Coast” construction practices. The second floor has both a home office and a laboratory environment area. Some nonstructural components on this level are anchored whereas others are unanchored for comparison. The third floor has a server room layout (Figure 7a). Sensitive electronic equipment is often damaged or nonfunctional after a major quake and the explicit focus of this floor is to investigate the performance of computer equipment in such a setting. Since essential facilities such as hospitals have a critical need to be operational after an earthquake, levels four and five are designated as hospital floors. The fourth floor represents an intensive care unit (ICU) layout and the fifth floor represents a surgery suite. The ICU and surgery suite are outfitted with large medical equipment typical to those rooms. The reason for having different architectural floor plans on each floor was to observe the behavior of nonstructural components and systems during ground motions in a wide variety of occupancy scenarios and to take advantage of the space available. The installation processes, design processes, materials, anchorage, and other features of the nonstructural components are as close to actual field conditions as possible. If a variety of materials or attachments are used in practice, the most common are adopted, with some compliant with current code and others not. Different materials or products are used in some cases to facilitate comparison between different options used in practice. Industry partners designing the NCSs, including anchorage, were each given a nonstructural design guide, with loads and deformation distributions determined by the authors based on numerical simulations of the test building’s response. This supports consistent design amongst the product suppliers, while allowing for a more realistic design demand than might otherwise be obtained with the building design code. Mechanical-Electrical-Plumbing (MEP) Subsystems A partial mechanical-electrical-plumbing subsystem is installed in the building. Examples of the MEP subsystems include plumbing to support storm drain services, an electrical distribution system to support operation of the medical equipment, elevator, and lighting to name a few, and select levels of mechanical ducting to mimic components installed in a functioning HVAC system. The plumbing scope includes a 3” no-hub cast iron pipe that extends from a storm drain on the roof to the first level. Extending from 1’ below the second floor ceiling to 1’ above the fourth floor slab is a 4” steel pipe and a 1 ½” copper pipe. The location for these pipes was chosen such that they extended through the floors that expected the maximum interstory drifts. Similar strategies for the mechanical and electrical systems are implemented (Figure 7b). The electrical scope includes outlets at different locations on every level corresponding to the placement of equipment in

need of a power supply. In addition, fire sprinklers are installed at all levels of the building.

(a) (b) Figure 7. Level 3 floor plans: (a) architectural floor plan with IT server occupancy and burn floor and (b) MEP plan. Largest NCSs The largest NCSs installed in the test building include: (1) a functioning passenger elevator, (2) prefabricated metal stairs, (3) ceiling subsystems, (4) two pieces of roof mounted equipment and a penthouse, (5) three levels of balloon framed metal studs overlaid with an exterior insulation finishing system (EIFS) and (6) two levels of precast concrete cladding. The full-scale, fully functional elevator has a 3500-pound capacity, travels all five floors of the building for a total of 56’ in distance (Figure 8a). Its design is compact, with no need for a mechanical penthouse as the controller is integrated with the cab. The type of elevator installed has yet to be exposed to the seismically active regions in the U.S. (its only market is in Europe). An elevator is an essential component in hospitals and other emergency facilities, so it is important to understand the exact performance of it and how it interacts with the rest of the structure during a strong ground motion in order to find better methods to protect it and ensure usability after a major quake. Typical seismic damage to an elevator includes guide rail anchorage damage, bent guide rails, counterweight derailment, a jumped or twisted rope, the collision between the counterweight and car, etc. Likewise, the prefabricated metal stairs are installed to support access to all floors including the roof (Figure 8b). It is also essential that stairs remain operational after a large earthquake in any building. Damage to the stairs during an earthquake could potentially prevent evacuation and immediate operation. The stair system installed in the test building was previously tested by Higgins (2009) at full-scale using a onelevel configuration with hydraulic actuators at the ends.

(a) (b) Figure 8. NCSs providing building floor-floor access: (a) elevator rendering and (b) photograph of the prefabricated metal stairs in the test building. Different ceiling types including a “typical East Coast” design, code compliant design for seismic design categories D, E & F, designs per Seismic RX recommendations, a California Department of State Architect (DSA) drywall system, typical Office of Statewide Health Planning and Development (OSHPD) approved ceiling design for hospitals, and a new ceiling design are implemented on different levels of the building. The rooftop equipment consists of a stair penthouse, an air handling unit, and a roof mounted cooling tower (Figure 9a, 10). The air handling unit is directly mounted to the southwest corner of the roof using a typical anchorage design. The cooling tower is located at the northeast corner of the roof, and sits on a steel frame supported by vibration isolators. The isolators are bolted directly to four load cells, which are mounted to the roof slab and measure forces in five degrees of freedom. There are two types of exterior finishes on the building (Figure 9b, 10). Levels one through three incorporate a metal stud balloon framing system, while level four and five utilize a precast concrete panel system. The balloon framing system consists of continuous metal studs extending through three levels, with king studs framing the openings. The metal studs are connected with a clip connection for lateral support at each floor level. The first level has openings that are governed by the need to post tension the vertical rods connecting the foundation to the shake table for the fixed base testing condition.

(a) (b) Figure 9. Photographs of (a) roof mounted NCSs and (b) installation of level 1-3 balloon framing at exterior.

The exterior cladding on the fourth and fifth floors consists of eight precast concrete panels configured in a punch-out window floor-floor configuration. Steel embeds supporting connection of these panels to the structure were cast with the slab, columns and beams as would be conventionally done in practice. By using eight panels per floor, different types of connection configurations are utilized for comparative purposes. The system is designed to withstand an ultimate interstory drift of 3.5%. Push-pull connections are used to allow the panels to move with the structure, while bearing connections are utilized to support the panels at the deck level.

(a) (b) Figure 10. 3D rendering of structure (a) view from northeast corner and (b) view from southwest corner. Instrumentation Approximately 500 accelerometer, displacement, and strain sensors are planned for deployment on the structure, the majority of these are dedicated to measuring the response of the NCSs. Accelerometers are used to monitor the behavior of the structure, including the displacement, velocity and acceleration time history in three different directions and at every floor level. In addition, a GPS system will be implemented, and this data will be compared with the displacements found from the double-integration of the acceleration reading recorded by the accelerometers. Displacement potentiometers will be placed on the structure to measure the plastic hinge rotations, the shear deformation of the joints, and elongations in the slab and beams. Relative motion between the structural and nonstructural components will be captured using the analog sensors. Both the structure and the NCSs will be monitored with an array of nearly 70 digital cameras.

Test Protocol The building will first be tested in the base isolated configuration and subjected to three near-serviceability level motions, two of which will be repeated when the building is fixed to its base. White noise testing will be conducted before and after each motion to characterize the state of the structure and the nonstructural systems. The building will then be tested in the fixed base condition using serviceability and a scaled MCE level motion. The scaled MCE motion will be implemented at the end of testing and correspond to a performance target of approximately 2-2.5% interstory drift. Following seismic shaking, local burn (fire) tests will be conducted on designated areas of the third floor (Figure 7a). The areas where this will occur are sealed off with typical code-compliant fireproofing prior to the shake testing. After the earthquake simulations, controlled smoke will be released in the places of interest, and the propagation of the smoke through any damaged portions of the building will be monitored and recorded. Concluding Remarks A full scale, five-story cast-in-place reinforced concrete test building has been constructed on the largest outdoor shake table in the United States. Unique to prior test programs, it is fully outfitted with a wide variety of typical nonstructural components and systems that represent different occupancy/risk categories. The building will be shake table tested in two different configurations: first with base isolators and then with a fixed base. After shake table testing is completed, burn tests will be conducted in the structure, to observe the propagation of smoke through a seismically damaged building, representing the very dangerous probable threat of post-earthquake fires. This project will lead to breakthrough advances in the understanding of the complete structural and nonstructural responses during earthquakes and post-earthquake fires. Acknowledgements This project is a collaboration between four academic institutions (University of California, San Diego, San Diego State University, Howard University, and Worcester Polytechnic Institute), four government or granting agencies (the National Science Foundation, the Englekirk Advisory Board, the Charles Pankow Foundation, and the California Seismic Safety Commission), over 30 industry partners, and two oversight committees. A listing of industry project sponsors may be found on the project website: http://bncs.ucsd.edu/index.html. Through the NSF-NEESR program, funding is provided by grant number CMMI-0936505, where Dr. Joy Pauschke is the program manager. The above continuous support is gratefully acknowledged. In addition, the technical support of NEES@UCSD staff, and consulting contributions from Robert Bachman, chair of the project's Engineering Regulatory Committee, are greatly appreciated. Opinions and findings of this study are of the authors and do not necessarily reflect those of the sponsors.

References BNCS Project Website. “Full-Scale Structural and Nonstructural Building System Performance During Earthquakes & Post-Earthquake Fire”. [Dec. 14, 2011]. Chang, B.J., Hutchinson, T.C., and Englekirk, R.E. (2008). "Experimental seismic performance evaluation of innovative beam-column subassemblies." Structural. Systems Research Project Report Series. SSRP 08/01. Dao, N., Ryan, K., Sato, E., Okazaki, T., Mahin, S., Zahgi A., Kajiwara, K., and Matsumori, T. (2011). “Experimental Evaluation of an Innovative Isolation System for a Lightweight Steel Moment Frame Building at E-Defense.” ASCE Structures Congress. Mar. 29-31. Chicago, Illinois. Higgins, C. (2009) “Prefabricated Steel Stair Performance Under Combined Seismic and Gravity Loads”. Journal of Structural Engineering. ASCE,122-129. Hoehler, M., Panagiotou, M., Restrepo, J., Silva, J., Floriani, L., Bourgund, U., and Gassner, H. (2009). “Performance of Suspended Pipes and Their Anchorages During Shake Table Testing of a Seven-Story Building.” Earthquake Spectra, Vol. 25 issue 1: 71-91. Hutchinson, T., Hoehler, M., Gastelum, A., Watkins, D., and Wood, R. (2010), “Observations from the Sierra El Mayor earthquake (Northern Baja California Earthquake).” EERI: Report on Nonstructural Damage, Baja California Earthquake, Magnitude 7.2 on April 4, 2010. Matsuoka, Y., Suita, K., Yamada, S., Shimada, Y. and Akazawa, M. (2008), “NonStructural Component Performance in 4-Story Frame Tested to Collapse.” The 14th World Conference on Earthquake Engineering. Oct. 12-17, 2008. Beijing, China Miranda, E. and Taghavi, S. (2003). “Response Assessment of Nonstructural Building Elements.” PEER Report 2003/05. 64-65. Moehle, J., Riddell, R., Boroschek, R. et al. (2010), “Learning from Earthquakes: The Mw 8.8 Chile Earthquake of February 27, 2010.” EERI Special Earthquake Report. Mosqueda, G., Retamales, R., Filiatrault, A. and Reinhorn, A. (2009), "Testing facility for experimental evaluation of non-structural components under fullscale floor motions." The Structural Design of Tall and Special Buildings, 18: 387 NEES@UCSD. Network for Earthquake Engineering at the University of California, San Diego. [Dec. 14, 2011]. Sato, E., Furukawa, S., Kakehi, A. and Nakashima, M. (2011), "Full-scale shaking table test for examination of safety and functionality of base-isolated medical facilities." Earthquake Engineering & Structural Dynamics, 40: 1435–1453. Warcholik, G. and Priestley, M. (1998a). "Structural systems research project: High strength concrete joints tests." Report TR-98/01. Warcholik, G. and Priestley, M. (1998b). “Structural systems research project: High strength concrete joints tests - series 3." Report TR-98/12.