Embed Size (px)
Citation preview
Consortium of Universities for Research in Earthquake Engineering1301 S. 46th Street, Richmond, CA 94804-4698 tel: 510-231-9557 fax: 510-231-5664
http://www.curee.orgCUREE
The Establishment of NEESadapted from the 2002 CUREE Calendar
Introduction by Robert Reitherman
© 2001 CUREE. All rights reserved.
1
The Establishment of the Network for Earthquake Engineering SimulationPast CUREE calendars have sometimes contained short illustrated essays on the history of earthquake engineering. This year’s calendar concerns adevelopment that has yet to fully take shape but which may be looked back upon as one of the most important developments in earthquake engineeringin the twenty-first century: the establishment of the George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES).
Following a multi-year planning effort in the 1990s, the National Science Foundationobtained an $82 million budget approval from Congress for 2000-2004 for the firstMajor Research Equipment program in NSF’s Engineering Directorate. Named afterthe late Congressman Brown, who was a strong advocate of science and engineeringresearch, the NEES program has its home at NSF within the Division of Civil andMechanical Systems, a department within the Engineering Directorate, one of sevendirectorates at NSF.
The NEES Collaboratory can be visualized as a single laboratory containing next-generation experimental equipment and instrumentation, with an advanced informa-tion technology infrastructure for curated data repositories, teleparticipation, andsimulation. This “laboratory” where collaborative work is conducted will actually bea set of laboratories with wide geographic distribution, and with researchers, the usersof research, and computers that are not necessarily co-located with the researchfacilities involved in a particular project. Furthermore, the investment in experimentalfacilities is intended to improve simulation-based modeling applicable to a wide varietyof cases, rather than to build up a database of experimental results for application onlyto conditions very similar to those tested in a laboratory.
The term “NEES Collaboratory” encompasses more than the three NSF-fundedprogram components described below: System Integration, Equipment Sites, and
Consortium Development. According to National Collaboratories: Applying InformationTechnology for Scientific Research, a collaboratory is a network-enabled “...‘centerwithout walls’ in which the nation’s researchers can perform their research without regardto geographical location, interacting with colleagues, accessing instrumentation, sharingdata and computational resources, and accessing information in digital libraries.” (Na-tional Research Council, 1993. National Collaboratories: Applying Information Technol-ogy for Scientific Research. Washington, DC: National Academy Press.) Collaboratorieshave been successfully created in other fields, such as astronomy, where scientists atvarious locales share the use and control of unique observatory equipment. Collaboratorieshave also been established in physics and medicine. For example, a researcher in one city
NEES GLOSSARY
NEES: The George E. Brown, Jr. Network for EarthquakeEngineering Simulation, a Major Research Equipment (MRE) programof the Engineering Directorate of the National Science Foundation.
NEES Collaboratory: the sum of the components of NEES as theyoperate in the research mode, scheduled to be fully operational by2004 and operate for a decade.
System Integration: Development of the network, services, curatedrepository, and applications that tie the distributed experimental sitesand researchers together, scheduled to be completed by 2004.
NEESgrid: The Grid-based architecture selected by the SystemIntegration team for the network links, services and applications to beused by the NEES Collaborator.
NEES Equipment Sites: Engineering laboratories funded by NSF tocreate or enhance experimental facilities to be used in the NEESCollaboratory. Other sites are also expected to participate in theCollaboratory.
NEES Consortium: the non-profit entity that will manage the NEESCollaboratory.
“A collaboratory is a network-enabled centerwithout walls...researchers can interact withcolleagues...accessing instrumentation...sharing data, computational resources, anddigital libraries...”
2
can teleoperate an electron microscope located in another, and simulations requiring coordination of networkedcomputing power have been co-managed and co-operated by a team of researchers from several different institutions.NEES is the first collaboratory in the civil engineering field, and thus there are high expectations for it to illustratea model for a new style of research that will confer benefits not only on earthquake engineering research and earthquakeloss reduction but on the wider engineering realm. It is likely that within a few years professional engineers willpractice quite differently than is done today. NEES can be the leading edge for facilitating the transformation ofengineering practice in an era of information technology advancements.
System Integration awards in two phases totalling ap-proximately $10 million have been made by NSF to theUniversity of Illinois at Urbana-Champaign, under thedirection of Principal Investigator Thomas Prudhommeof the National Center for Supercomputing Applications. The System Integration project willcomplete the establishment of the NEES information technology infrastructure in 2004.NEESgrid will link laboratories, computational resources, researchers and research usersaround the country. Other experimental or simulation facilities will be able to network andbenefit from NEES Collaboratory interaction, and it is also being planned with world-wideaccessibility in mind. NEESgrid is the essential technical capability that enables NEES tofunction as a collaboratory, rather than as a collection of independently operated facilities andresearchers. See the “January” page of this calendar for more information on SystemIntegration.
NSF has also overseenthe first phase of a com-petitive proposal processwhereby 11 awards to-talling $45 million havebeen made for ShakeT a b l e , T s u n a m i ,
Geotechnical Centrifuge, Field Investigation, and Large-Scale Structural Testingequipment. Each of these eleven Equipment Sites is featured on one of the monthpages of this calendar. A directory of these facilities and connections to their websitescan also be found at the nees.org website. The second phase of awards is currentlythe subject of an NSF Solicitation, NSF 01-164, NEES Earthquake EngineeringResearch Equipment, Phase 2.
NSF’S VISION FOR NEES
“Through NEES, the earthquakeengineering community will useadvanced experimental capabilitiesto test and validate more complexand comprehensive analytical andcomputer numerical models thatwill improve the seismic designand performance of our Nation’scivil and mechanical systems.”
NSF 01-56NEES does not seek incremental improve-ment of earthquake engineering research:It seeks to revolutionize it. NEES will bringmany new things to the earthquake engineeringfield:
-- curated data repositories-- teleobservation and teleparticipation-- data visualization-- real-time multi-lab interaction-- a national computational infrastructure-- sharing of facilities-- rapid sharing of data and models
If it sounds different, it’s becauseit hasn’t been done before.
First Phase NEES Equipment Site Awardees
Shake Tables: University at Buffalo-SUNY; University of Nevada at RenoCentrifuges: Rensselaer Polytechnic Institute; University of California, DavisTsunami Wave Basin: Oregon State UniversityLarge-Scale Laboratory Experimentation: University at Buffalo-SUNY; University of Minnesota; University of California, Berkeley; University of Colorado at Boulder;Field Experimentation: University of Texas at Austin; University of California, Los Angeles
For more information go to: www.nees.org
3
The third leg of NSF’s NEES establishment effort is also now under construction: the formation of the NEES Consortium that will operate and manageNEES for its initial decade of planned operation, from the fall of 2004 through 2014. This award has been made to the Consortium of Universities forResearch in Earthquake Engineering (CUREE). The Earthquake Engineering Research Institute and the Civil Engineering Research Foundation of theAmerican Society of Civil Engineers are partners in the project. The entire earthquake engineering community is afforded an opportunity to becomeinvolved in shaping the vision of NEES Collaboratory research and specifically in having a voice in setting the new organization’s policies, governancestructure, procedures -- in setting its fundamental character. This community extends beyond researchers who have been active in the earthquakeengineering field in the past to also embrace practitioners, agencies with earthquake-related responsibilities, educators, and others who have a stake inthis venture.
Task Group onInformation Technology
and Facilities
Cherri Pancake, Co-PIOregon State [email protected]
Catherine FrenchUniversity of Minnesota
Anke KamrathUCSD Supercomputer Center
Task Group onCommunity Development
Sharon Wood, Co-PIUniversity of [email protected]
Thalia AnagnosSan Jose State [email protected]
Dante FrattaLouisiana State University
J. Peter KissingerCERF - ASCE
Kim RoddisUniversity of [email protected]
Task Group onCollaboratory Research
Stephen Mahin, Co-PIUniversity of California-Berkeley
Jeremy IsenbergWeidlinger Associates Inc.
Philip LiuCornell [email protected]
David McCallenLawrence Livermore National Lab
Thomas O’RourkeCornell [email protected]
Task Group onConsortium Organization
Robert NigborUniversity of Southern California
William HolmesRutherford & Chekene
Susan TubbesingEERI
Gene WhitneyOffice of Science & Technology
Policy (OSTP)[email protected]
Joy Pauschke Program Director, [email protected]
Robert Reitherman, Principal Investigator,CUREE - [email protected]
Thomas Finholt, CREW, University of Michigan,Project Assessment - [email protected]
Andrew Neitlich, The Sago Group, OrganizationalDevelopment Consultant - [email protected]
NEES Consortium Development Project
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
4
A consortium of institutions led by the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign is building NEESgrid, a central component of a national virtual earthquake engineering laboratory,
or “collaboratory,” supported by $10 million in funding from the National Science Foundation. The effort is part of the NSF’sGeorge E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES) project. The integrated NEES network, calledNEESgrid, will link earthquake engineering research sites across the country, provide data storage facilities and repositories, andoffer remote access to the latest research tools and advanced experimental facilities located at leading universities around thecountry.
Development of NEESgrid will continue through Sept. 30, 2004. A community-based NEES Consortium will operate thecollaboratory beginning in October 2004. NEESgrid will take advantage of the grid tools and technologies developed over the lastfive years through the NSF’s Partnerships for Advanced Computational Infrastructure (PACI) program. Proven grid technolo-gies—such as the Globus Toolkit for distributed computing, developed by Argonne National Laboratory and the InformationSciences Institute (ISI) at the University of Southern California (USC), both partners in the effort—will be incorporated intoNEESgrid. Globus will allow researchers to seamlessly share experimental equipment, computational resources, and research data.
In addition, NEESgrid will include collaboration and teleoperation tools developed through the NSF’s Knowledge and DistributedIntelligence and Information Technology Research initiatives and through the Department of Energy’s DOE2000 effort. The
Collaboratory for Research on Electronic Work (CREW) at the University of Michigan’s School of Information and Argonne’s materials science division have developed operationalcollaboratories under these programs and are also members of the NEESgrid team. Joining NCSA, Argonne, ISI, and CREW in developing the NEESgrid are the UIUC and USC civil engineeringdepartments.
“The goal is to create a collaborative research network by linking researchers and engineering testing facilities across the United States, and providing them with the latest computational tools,”said Priscilla Nelson, NSF division director for Civil and Mechanical Systems. “We expect this network to speed the simulations, experiments, and data analysis that lead to better seismic designand hazard mitigation.”
“The NEES Collaboratory and the NEESgrid will take many of the technologies developed through the PACI program and refine them for use by a specific technical community,” accordingto Dan Reed, director of NCSA and the National Computational Science Alliance. “This is a way for us to show the impact of the work we’ve been doing for the last five years and how we canapply what we’ve learned to the real-life needs of a group of scientists and engineers.”
NEES will allow researchers to share and remotely operate experimental equipment at a number of earthquake engineering facilities linked to the network and to more easily share data andcomputations. The equipment includes shake tables, geotechnical centrifuges, a tsunami wave basin, large-scale structural testing, and field-deployed apparatus and instrumentation. The goalis to model and analyze earthquake effects and help engineers design buildings and infrastructure to withstand them.
“NEESgrid will be an environment not only for research engineers but for practicing engineers who are involved in the actual design and development of roads, bridges, dams, and buildings,”said Tom Prudhomme, NEESgrid Principal Investigator. “Practicing engineers don’t usually use research data and complex simulation models in their work because they don’t have easy accessto it or effective ways to validate the results. That is about to change.”
NEESgrid will support national collaborations among earth-quake engineering researchers, and provide easy access toadvanced computing resources.
NEESgrid: A Virtual CollaboratoryFor Earthquake Engineering Simulation
University of Illinois, Urbana-ChampaignThomas Prudhomme, Principal Investigator
Website: http://www.neesgrid.org/
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
5
Testing of full-scale structure on dual shake tables (100 metric tons) with alternative protective system
The University at Buffalo’s Structural Engineering and Earthquake Simulation Laboratory (SEESL), which is the flagship laboratory in the Multidisciplinary Center for EarthquakeEngineering Research (MCEER), will be an important node of a nationwide “collaboratory” in the NSF’s Network for Earthquake Engineering Simulation (NEES).
A key element of the upgrade of SEESL under NEES is the installation of two moveable, six degrees-of-freedom shake tables. The intent of the NEES Node project at the University at Buffalois to develop the most versatile earthquake engineering research facility possible, designed to provide testing capabilities that will revolutionize the understanding of how very large structuresreact to a wide range of seismic activity, even when tested to complete failure. Specifics of the installation include:
• A set of two high-performance, six degrees-of-freedom shake tables, which can be rapidly repositioned from directly adjacent to one another to positions up to 100 feet apart (center-to-center).
• Together, the tables can host specimens of up to 100 metric tons and as long as 120 feet, and subject them to fully in-phase or totally uncorrelated dynamic excitations.
• Next to the shake tables, a large reaction wall equipped with external dynamic actuators will allow application of computer-simulated forces on shake table- mounted substructures, thus simulating the reaction of the entire structure (Real-Time Dynamic Hybrid Testing, or RTDHT).
• Equipment required to operate the shake tables, including a high-capacity, high-performance hydraulic supply and distribution system (up to 1600 gpm), and numerous digital control systems.
• Networked tele-experimentation capabilities using modular and expandable teleobservation and teleoperation equipment, tied to the testing systems using discrete and global sensors, including high-resolution digital video and imag- ing capabilities.
• The SEESL facility will include a testing area expanded from approximately 3,000 square feet to 13,000 square feet (including a greatly expanded strong floor), reaction walls, and a trench for the moveable shake tables.
Versatile High Performance Shake Tables Facility TowardsReal-Time Hybrid Seismic Testing
University at Buffalo, SUNYMichel Bruneau, Principal Investigator
Website: http://civil.eng.buffalo.edu/seesl/
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
6
The high-bay Large-Scale Structures Laboratory (LSSL) at the University of Nevada, Reno was established in 1992 and equipped with two 450 kN shake tables funded by the Federal Emergency Management Agency in 1995. The building was expanded in 1999 to approximately 780 sq m.
The main test floor of the LSSL is a heavily reinforced concrete slab providing 780 sq m of useable test area. This area is serviced by two 222 kNoverhead cranes with a clear height of 11 m. The strong floor measures, 46 x 17 x 0.91 m (length x width x thickness), and has about 1325 tie downholes on a 0.61 x 0.61 m grid. Significant testing is conducted in the LSSL using independent actuators. For this purpose, the laboratory has severalMTS servo-controlled static as well as fatigue-rated actuators ranging in size from 250 kN to 3.1 MN, with stroke capacities ranging from ±75 mmto ±600 mm. These are routinely used for large-scale experiments on structural components that are too large or unsuitable for shake table execution.Hydraulic hardlines feed the Laboratory and provide 945 l/min at 21 MPa. Four ports along the centerline of the Laboratory floor provide accessto feed, return, and drain lines. The same distribution system can supply up to 1,890 l/min depending on the pump capacity. At present, theLaboratory’s pumping capacity is 1000 l/min, which is boosted up to 6,056 l/min with blowdown accumulators for short durations.
A major upgrade and expansion of the LSSL will be undertaken under the NEES Equipment Award from the National Science Foundation,supplemented by awards from the Department of Housing and Urban Development and the Department of Energy.
The equipment to be obtained will enable the upgrade of the two existing tables from uniaxial to biaxial motion, and thepurchase, installation, and commissioning of a third biaxial table. Manufactured by MTS, each table measures 4.35 x4.50 m, has a stroke of ±300 mm, and can reach a peak velocity of 1 m/sec and an acceleration of ±1g under the full 450kN payload in both directions. All three shake tables will be relocatable with the same characteristics in both directions.To achieve these maxima, banks of blowdown accumulators are used. Currently, LSSL has two such accumulators andwill purchase a third for the new table.
Together the three tables can host specimens up to 1.35 MN in total weight, and can be separated a minimum distanceof about 9 m up to a maximum of 36.5 m, centerline-to-centerline. Each table may be operated independently of the othertwo tables, in-phase with the other two tables thus forming a single large table, or differentially with the other two tablesfor the simulation of spatial variation effects in earthquake ground motions. As part of the expanded system (upgradedand new shake tables), the following equipment will also be obtained and installed: new hydraulic distribution lines,blowdown bank, upgraded hydraulic power supply (addition of a third 700 l/min pump), digital control system for threetables, expansion of data acquisition system, and hardware/software to accommodate teleparticipation and data storage.
The new facility will also be telecapable, in the sense that it will be equipped and connected to the university’s highbandwidth, Internet-2 network for remote participation of off-site researchers in real-time.
A schematic representation of the Multiple Shake Table Facility
View of one of the uniaxial shake tables
A Biaxial Multiple Shake Table Research Facility
University of Nevada, RenoIan Buckle, Principal Investigator
Website: http://bric.ce.unr.edu/nees/nees.htm
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
7
In the fall of 2000, University of California at Davis received a $4.6 million award, “A NEES GeotechnicalCentrifuge,” NSF-CMS 0086566.
All of the following upgrades of the existing centrifuge at UC Davis are to be in place by fall of 2004:
• Increase the centrifuge capacity from 40 to 80 g• Hundreds of networked advanced sensors• High resolution, high-speed digital cameras• 4 Degree-of-freedom gantry robot for in-flight inspection and construction - on-board tool rack pile driving, soil improvement, cone, vane, geophysical testing, stereo-video inspection, versatile tool interface for custom tools - to be fully operational in conjunction with shaking table tests• In-flight geophysical testing and tomography - S- and P-waves - Electromagnetic waves• Vertical-horizontal biaxial shaking table• Remote teleoperation and telecollaboration• Visualization facility with 3 m x 3 m power wall• New 400m2 sample preparation and office building
The Center for Geotechnical Modeling (CGM) operates the NEES Geotechnical Centrifuge Facility at UC Davis. The vision of the CGMis to capitalize on revolutions in instrumentation and information technology to produce unambiguous experimental data for comprehensiveassessment of model-based simulation theories. Other goals of the CGM are (1) to organize and promote multi-campus researchinvestigating complex geotechnical problems, (2) to disseminate research findings through conventional and web publications, (3) to traincurrent and future geotechnical engineering professionals and researchers and (4) to develop and operate the upgraded centrifuge facility.
Pictured above is the centrifuge rotunda at UC Davis, housing the 9-meter radius geotechnical centrifuge. The flexible shear beam containeris loaded on the bucket and the rigid container (clear sides) is on the floor. Onboard computers are housed in the cabinets on top of the arm.Data is transmitted from these computers through a fiberoptic slip ring above them to the control room nearby. The gearbox is located underthe center cone and the drive motor is below the arm on the left.
UC Davis personnel involved in the NEES project are Principal Investigator: Bruce L. Kutter; Co-Principal Investigators: Daniel W. Wilson,Ross W. Boulanger, Boris Jeremic, Steven A. Velinsky, S. Ben Yoo, Bernd Hamann, Kwan-Liu Ma; and J. Carlos Santamarina (GeorgiaInstitute of Technology).
A conceptual diagram of the onboard robot andtool rack, to be operational by 2004.
Geotechnical Centrifuge at the University of California, Davis
A Geotechnical Modeling Facility
University of California, DavisBruce Kutter, Principal Investigator
Website: http://cgm.engr.ucdavis.edu/NEES/
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
8
Rensselaer’s centrifuge was commissioned in 1989 and started conducting physical model simulations of soil and soil-structure systemssubjected to in-flight earthquake shaking in 1991. In this decade of successful operation, the facility has published results of about 360
earthquake-related model simulations, served as the basis for several Ph.D. theses at RPI; contributed to the research of RPI faculty and studentsas well as of dozens of visiting scholars and outside users from the US, Asia, Europe, and Latin America; and provided data and research resultsto many people and organizations around the world. This centrifuge earthquake research has been conducted with two existing one-dimensionalin-flight shakers, which can accommodate respectively 90 kg and 450 kg payloads.
The centrifuge machine installed at Rensselaer is an Acutronic Model 665-1 constructed to RPI’s specifications. It has an in-flight platformradius of 3.0 m and can test a payload of 1 ton at 100 g or 0.5 ton at 200 g, corresponding to linear dimension scaling factors of 100 or 200,respectively. A small in-flight computer controlled earthquake simulator (shaker), capable of shaking a payload of 90 kg, was designed andconstructed at RPI in December 1990; a laminar rectangular box simulating the shear beam boundary conditions for the soil in the free field duringshaking was finalized in March 1991. A larger in-flight shaker capable of shaking a payload of 450 kg has started operation in 1995.
In addition to the shakers and the laminar box, development for equipment and instrumentation of bothstatic and earthquake loading has included the design and construction of rigid wall containers of varioussizes, a high-speed data acquisition system, a video system for in-flight real time monitoring, and theacquisition of suitable miniature LVDT’s, accelerometers, and pore pressure transducers.
Currently we are in the process of acquiring several new pieces of equipment as part of RPI’s participation in the NEES (Network for Earthquake EngineeringSimulation) NSF project, which provides funding for enhancing the earthquake simulation capability and for networking several earthquake experimentalfacilities in the USA. The NEES next-generation earthquake engineering capability for the RPI centrifuge facility includes: (1) a 2D in-flight earthquakeshaker (two prototype horizontal components) and associated 2D laminar box container to allow for more realistic 2D modeling; (2) a four degrees of freedom(4D) robot capable of performing in flight operations such as construction and excavation, pile driving, ground remediation, cone penetration, and staticand cyclic loading tests without stopping the centrifuge; (3) a networked data acquisition system with Internet Teleobservation / Teleoperation capability,to be linked to the high-speed RPI Gigabit Ethernet Backbone; (4) two high-speed cameras and image processing software; (5) development of a newgeneration of advanced and improved sensors capable of providing a better resolution of the measured model response; and (6) other equipment aimed atincreasing the capability of the centrifuge to test a greater number and wider variety of earthquake engineering models.
Especially important in the vision embodied in the NEES project is the future use of dense arrays of advanced sensors and of high speed cameras to providehigh resolution measured model response. In conjunction with the proposed networked data acquisition system with remote access capability, this will allowfor a quantum jump in the use of the data at RPI and throughout NEES, including teleobservation, shared-used of data, test visualizations, system
identification, numerical computations and development of model-based simulations. In addition, the NEES equipment will allow for teleoperation/control over the Internet. The RPI NEESproject is managed by Professors R. Dobry, T. Abdoun, M. Zeghal and T. Zimmie from RPI and A. Elgamal from UCSD. This team covers all areas of expertise needed for a successful project,ranging from centrifuge model testing and equipment development to system identification, visualization, and computational simulations.
Upgrading, Development and Integration of Next GenerationEarthquake Engineering Experimental Capability
Rensselaer Polytechnic InstituteRicardo Dobry, Principal Investigator
Website: http://www.ce.rpi.edu/centrifuge
In-flight four degree-of-freedom robotat the LCPC centrifuge in France; asimilar in-flight robot will be installedat the RPI centrifuge.
One-dimensional in-flight shaker with 450 kgpayload capacity currently in operation at the RPIcentrifuge.
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
9
In 2000, the O. H. Hinsdale Wave Research Laboratory received a $4.8 million grant from the National Science Foundation (NSF) Networkfor Earthquake Engineering Simulation Program (NEES) to construct the world’s largest and most advanced tsunami testing facility in
support of national experimental and computer simulation research. Currently, the 3-D basin is being extended to 49.4 m (160 ft) long, 26.5m(87 ft) wide and 2 m (6.6 ft) deep with a 29-segment directional, spectral wave generator located along one of the 26.5 m (87 ft) walls. Eachsegment of the new wave generator will have a maximum stroke of 2 m (6.6 ft) and a maximum velocity of 2 m/s (6.6 ft/s). The WRL will havea comprehensive information architecture to support remote users and a tsunami experiment databank for the broader research community tostudy the results of tsunami experiments.
The O.H. Hinsdale Wave Research Laboratory originated in 1972 with a gift from a private benefactor to construct a long, two-dimensional(2-D) wave channel to examine the stability of artificial islands for offshore nuclear power plants. In 1986, the Office of Naval Researchprovided funding to construct a rectangular, directional wave basin to examine diffractive and refractive wave processes in a real, three-dimensional (3-D) environment. These two facilities, housed in a 1.15 acre steel building, are particularly suitable for tsunami research.
2-D Wave Channel — The large 2-D wave channel is 104 m (342 ft) long, 3.66 m (12 ft) wide and 4.57 m (15 ft) deep. A hinged flap waveboard, hydraulically driven and direct digital controlled, generates periodic and random waves up to 1.52 m (5 ft) high. The full range of
wavelengths can be modeled from short, deepwater waves tolong, shallow water waves including solitary waves andtsunamis. The actively controlled wave generator simulatesthe open boundary condition at sea, minimizes reflection and maintains high quality incident wave environment.This wave channel, which is the largest and most technically advanced 2-D facility of any university, worldwide,is used to induce and measure wave forces on constructed and natural, ocean features at large Reynolds numbers,providing high-resolution wave structure interactions with minimum distortion due to viscous effects. Exampleapplications, in addition to seismically induced tsunami and solitary waves, include wave forces on marinepipelines and piles, stability of seabeds and breakwaters, runup on coastal margins and floating or fixed structureresponse to wave excitation.
3-D Directional Wave Basin — The existing directional wave basin is 26.5 m (87 ft) long, 18.3 m (60 ft) wideand 1.52 m (5 ft) deep, with a 30-segment directional, spectral wave generator located along one of the 18.3 m
(60 ft) walls. The 0.61 m (2.0 ft)-wide wave generator segments, powered by direct digital controlled AC motors at the edge of each segment, provide a continuous wave front and minimizelateral spurious wave activity. The actively controlled wave boards can generate incident wave heights up to 0.67 m (2.2 ft) at peak spectral periods of 2-3 seconds. The wave-board motionis of the piston variety, yielding excellent intermediate to long wave behavior. By changing the relative phase of each wave maker segment, one can control the time at which a wavelet leavesthe wave board, thereby generating waves that propagate at an angle to the face of the in-line wave generators. This 3-D facility complements the 2-D wave channel by providing the abilityto model structures that are sensitive to the orientation of the incident wave and structures that are large enough to change the wave direction. Example applications, in addition to seismicallyinduced tsunami and solitary waves, include directional wave effects on ship motions, gravity oil production platforms, artificial islands, floating bridges, jetty headlands, and port facilities.
OSU Two-dimensional Wave Channel
OSU Three-Dimensional Wave Basin
Multidirectional Wave Basin for Remote Tsunami Research
Oregon State UniversitySolomon Yim, Principal Investigator
Website: http://www.nees.orst.edu
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
10
Cable-stayed bridge segment with RTDHT using two shake tables, reaction walls, andlarge-scale high performance actuators
The University at Buffalo’s Structural Engineering and Earthquake Simulation Laboratory (SEESL),which is the flagship laboratory in the Multidisciplinary Center for Earthquake Engineering
Research (MCEER), will be an important node of a nationwide “collaboratory” in the NSF’s Networkfor Earthquake Engineering Simulation (NEES).
Key elements of the upgrade of SEESL under NEES include new reaction walls, significant enlarge-ment of the strong floor area, dynamic and static actuators, and associated control systems – allintegrated into a new dual shake table facility. The intent of the NEES Node project at UB is to developthe most versatile earthquake engineering research facility possible, designed to provide testingcapabilities that will revolutionize the understanding of how very large structures react to a wide rangeof seismic activity, even when tested to complete failure. The facility will be capable of conductingtesting of full or large-scale structures using static or dynamic loading. The use of modern techniquessuch as Pseudo-Dynamic, Effective Force, and Real-Time Dynamic/Pseudo-Dynamic Hybrid will bepossible, along with conventional Static, Quasi-static, and Dynamic Force techniques. Real-TimeDynamic Hybrid Testing (RTDHT) is a new form of testing being explored at UB. The implementationincludes shake table and dynamic force experiments of substructures, combined with real-timecomputer simulations to provide the most complete picture of how earthquakes affect large buildings,bridges and other structures. Specifics elements include:
• Reaction walls of 12.9 m by 13.5 m, strong floor of 25.8 m x 13.7 m, and large-scale, high-performance dynamic and static actuators, providing the ability to conduct dynamic, pseudo-dynamic, hybrid pseudo-dynamic testing.
• The actuators and associated control systems will allow for the development of new testing methodologies, including the effective force control testing method, in which large structures could be directly subjected to dynamic excitations without the need for shake tables.
• The large reaction wall, built next to the shake tables and equipped with dynamic actuators, will allow for the application of computer-predicted external forces on shake table-mounted substructures, thus simulating the reaction of the entire structure (Real-Time Dynamic Hybrid Testing, or RTDHT).
• Equipment required to operate the actuators, including a high-capacity, high-performance hydraulic supply and distribution system (up to 1600 gpm), and numerous digital control systems.
• Networked tele-experimentation capabilities using modular and expandable teleobservation and teleoperation equipment, tied to the testing systems using discrete and global sensors, including high-resolution digital video and imaging capabilities.
Large-Scale High Performance Testing Facility TowardsReal-Time Hybrid Seismic Testing
University at Buffalo, SUNYMichel Bruneau, Principal Investigator
Website: http://civil.eng.buffalo.edu/seesl/
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
11
The Reconfigurable Reaction Wall-Based Earthquake Simulation Facility (RRW ESF) is designed to support the developmentof a new generation of hybrid testing methods that smoothly integrate physical and numerical simulations. These methods are
based on the concept of sub-structuring: Portions of the structure expected to behave in a predictable manner are modeled numerically,while one or more complex sub-assemblies and the boundaries on which they interact are modeled using scaled physical models. Usingnumerical integration algorithms, the physical and numerical sub-structures can be analyzed as a single structure. Such hybrid testingmethods, better know as variations of the pseudo-dynamic testing method, have been pioneered at Berkeley in the early 1980s and weresubsequently adopted world-wide. Recent advances in digital control theory, as well as exponential increases in numerical processingpower and speed of inter-processor communication, make it possible to develop a new generation of hybrid simulation methods, thereal-time multiply-substructured pseudo-dynamic testing methods (RT MS-PDTM). The sub-structures, physical or numerical,involved in an MS-PDTM test need not be at the same geographic location. Instead, the network connecting them becomes thesimulation. Enabling such distributed real-time simulation is the core of NEES, making the proposed RRW ESF one of its key elements.
The RRW ESF leverages the capabilities of an existing testing facility located at theUniversity of California Richmond Field Station. The reconfigurable reaction wall will bebuilt across the width of an existing 80-by-20-foot strong floor. It comprises twelve 3-foot-tall reinforced concrete hollow blocks that can be post-tensioned together in threeconfigurations (Figure 1). In the first configuration (Figure 2), a 42-foot wall can be erectedin the middle of the floor to enable simultaneous pseudo-dynamic testing of two sub-structures. In the second configuration (Figure 3), the full-height wall can be positioned
at the edge of the strong floor, next to an existing 4 million pound axial compression-tension machine. This configuration makes it possible to testsub-structures up to 33 feet tall under various combinations of bending, shear, and very high axial forces, creating a unique asset in NEES facilitiesportfolio. Finally, the third wall configuration (Figure 4) comprises two 21 foot tall walls positioned at the ends of the strong floor. This “two-policeman” configuration enables testing of collapse mechanisms in planar substructures where axial forces and deformations in beams may play animportant role. A digital servo-control system for hybrid test control and a 128-channel data acquisition system are the electronic backbone of theRRW ESF. They will be directly connected to the NEES network to provide access to parallel computing facilities hosting the numerical sub-modelsin NEES simulations. The heart of the RRW ESF is the combination of an existing hydraulic pump system and a bank of accumulators. They willpower an array of static and dynamic actuators, spanning a force range between 25 and 750 kips and a stroke range between 12 and 72 inches. Thisactuator array is designed to utilize the full reaction and drift potential of the proposed reconfigurable reaction wall facility.
The RRW ESF is a capable stand-alone facility. However, it becomes even more potent when networked together with other NEES experimental andcomputational facilities. Working together in a network, these facilities will make simulations more real, improve their economy, and significantly enlarge the range of structures and systemswe can simulate before we build them. Earthquake engineers sorely need such simulations to improve seismic performance of our infrastructure, reduce economic losses in natural disasters,and, most importantly, make structures safer and more reliable.
Figure 1
Figure 2
Figure 4
Figure 3
Reconfigurable Reaction Wall-Based EarthquakeSimulator Facility
University of California, BerkeleyJack Moehle, Principal Investigator
Website: http://nees.berkeley.edu/
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
12
A Fast Hybrid Test (FHT) System is currently under development at the University of Colorado, Boulder as part of the NEES program. The system will allow efficient and realistic evaluation of the performance of structural systems and components under earthquake loads. The FHT system is based on the pseudodynamic test concept, and it combines physical testing with model-
based simulation. The main distinction is, however, that the system is able to achieve a rate of loading that is significantly higher than that in a conventional pseudodynamic test, approachingthe real-time response of a structure under earthquake loads. In such a test, the hydraulic actuators will move continuously based on command signals generated by a closed-loop feedback andnumerical computations. The system will incorporate a substructuring capability, whereby only a portion of a structure can be tested and the rest of it is simulated by a finite element modelusing nonlinear beam-column elements. With this, one can test the most critical structural subassemblage, where severe inelastic deformation is expected to develop, and model the rest of thestructure in a computer to provide a realistic boundary condition for testing. Furthermore, via a high-performance information network, the FHT System can be used to link geographicallydistributed laboratories to test different structural subassemblages and components of a large structural system using a single simulation model. Similar to other NEES sites, the system willbe connected to a high-performance network via Gigabit Ethernet to allow teleparticipation by remote users.
Envisioned Experiments
The FHT system can be used to support a variety of testing needs. Two examples are described below to illustrate its potential use.
As shown in Fig. 1, the FHT method is especially useful for evaluating the performance of shear walls. With this method, one canonly test the bottom story wall where most of the damage is expected to concentrate in a multi-story building, and model the restof the building using a mathematical model. The wall specimen in such a test will be controlled by three actuators to simulate theresponse under in-plane earthquake loads. One actuator is to control the horizontal displacement, and the other two to control thevertical displacement and rotation at the top.
During an earthquake, long bridge structures can be subject to multiple-support excitations. This can be simulated in a laboratoryusing multiple shake tables. The FHT method provides an efficient alternative to multiple-shake-table tests. One can test individualbridge columns and model the rest of the bridge structure using the FHT method. In such a case, asynchronous ground excitationsare generated by the mathematical model.
Major Equipment
The system will be housed in the Structures Laboratory of the University of Colorado that has a high bay with 2800 square feet of strong floor for testing structural components up to 10 ft.(3 m) tall. It will consist of three high-performance dynamic actuators and a digital servo-control system that has three control channels and a 48-channel data acquisition unit with signalconditioning. One actuator will have a 220 kip (100 ton) load capacity in tension and compression with a ± 6 in. (152 mm) stroke. It will be equipped with a 250 gpm (947 lpm) three-stagevalve. The other two will have a 110 kip (50 ton) load capacity and ± 5 in. (12 mm) stroke. Each of theses actuators will be equipped with a 180 gpm (682 lpm) valve.
The system will be powered by one 140 gpm (530 lpm) and two 70 gpm (265 lpm) hydraulic pumps that are already in place. Accumulator banks will be installed to provide supplemental oilflow for high-speed testing with seismic waveforms.
Figure 1: Shear Wall Tests
Fast Hybrid Test Platform for the Seismic PerformanceEvaluation of Structural Systems
University of Colorado at BoulderP. Benson Shing, Principal Investigator
Website: http://civil.colorado.edu/nees
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
13
The University of Minnesota Multi-Axial Subassemblage Testing (MAST) system, to be housed in a new laboratory on the Minneapolis campus, is one of four large-scale structural testingfacilities awarded through the National Science Foundation George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES) program. The MAST system enables multi-axial
cyclic static tests of large-scale structural subassemblages, including portions of beam-column frame systems, walls, and bridge piers. Six-degree-of-freedom control technology, employed bythe MAST system, advances the current state of technology in which case boundary effects are often reduced to simple uniaxial loading configurations because of difficulties encountered inimposing multiple-degree-of-freedom states of deformation and load using conventional structural testing means. The system is unique in size and scope and will greatly expand the large-scaleearthquake experimentation capabilities both nationally and internationally.
A rendering of the new laboratory and MAST system is shown schematically in the adjacent figure. The MASTsystem employs a stiff steel crosshead in the shape of a cruciform, which will be controlled with the six-degree-of-freedom control technology. Four ±1470 kN (±330 kip) vertical actuators, capable of applying a total force of±5870 kN (±1320 kips) with strokes of ±510 mm (±20 in.) will mount between the crosshead and the strong floor.Two sets of actuator pairs with strokes of ±400 mm (±16 in.), will provide lateral loads up to ±3910 kN (±880 kips)in the horizontal orthogonal directions. The actuator pairs will be secured to an L-shaped strong wall with universaltype swivels. The vertical clear distance permits specimens up to 7.6 m (25 ft.) in height to be tested. The horizontalclear distance between the vertical actuators can accommodate specimens up to 6.1 m (20 ft.) in length in the twoprimary orthogonal directions. Larger specimens may be oriented along the diagonal directions.
The stiff steel crosshead and six-degree-of-freedom control technology enable control of the position of a planein space. This feature makes it possible to apply pure planar translations, as well as the possibility of applyinggradients to simulate overturning (e.g., axial load gradient in the columns of a multi-bay frame, or wall rocking).Any degree-of-freedom may be programmed in either displacement control or load control, and degrees-of-freedom may be constrained in a master-slave relation to be a linear combination of the values of other degrees-of-freedom. As an example, using the mixed-mode control capabilities of the MAST, it is possible to program anylateral displacement history, and at the same time specify overturning moment as a constant times the lateral force,while simultaneously maintaining an independent history of axial load on the test specimen. The system will alsobe equipped with four ±980 kN (±220 kip) ancillary actuators with strokes of ±400 mm (±16 in.) which can be usedto apply lateral loads at intermediate story levels, gravity loads, or simulated specimen boundary conditions.
In conjunction with the contributions of the NEES System Integrator, the teleobservation/teleoperation infrastructure will provide all relevant information needed for both monitoring andinterpretation of the experiments. As such, this facility will incorporate real-time teleobservation and teleoperation of all visual monitoring information during an experimental run and real-timetransmittal of all acquired sensor data. Limited real-time teleoperation of hydraulic equipment will also be developed.
It is envisioned that the MAST facility will fit into an integrated data-centric approach for experimentation, computation, theory, databases, and model-based simulation facilitated through NEESSystem Integration. One of the most powerful features of this integrated approach to model-based simulation is an accumulated database of experimental results, which will feature the abilityto “replay” tests and to couple experimental responses with computer simulations.
Perspective view of MAST
A System for Multi-Axial Subassemblage Testing (MAST)
University of Minnesota-Twin CitiesCatherine French, Principal Investigator
Website: http://www.ce.umn.edu/mast
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
14
Previous studies of the seismic performance of full-scale structural systems have traditionally been extremely limited in their coupling of detailed performance data with nonlinear response,and this lack of data impacts design procedures in the form of significant uncertainty on the response of highly nonlinear structural systems. Since engineers cannot always wait for earthquakes
to provide excitation of structures from which measurements are needed, the vision of the University of California, Los Angeles equipment NEES site is to fill this critical gap in engineeringcharacterization of structural and geotechnical performance. To achieve this objective, this project seeks to develop and implement the next generation of forced-vibration testing and seismicstructural monitoring equipment.
Three categories of equipment are being developed for this mobile field laboratory: (1) equipment for detailed forced vibration testing of structural systems, (2) equipment and sensors for efficient,rapid evaluation of structural (i.e, acceleration, displacement, strain) and soil conditions (i.e., consistency, density, shear wave velocity, pore pressures, etc.) and for deployment of in situ sensorsto monitor soil response during forced and ambient vibration testing of structural systems, and (3) equipment for rapid data acquisition, processing, and distribution to enable near real-timeteleobservation and teleoperation.
A total of four vibration sources will be fabricated to enable flexible loading capabilities. Three eccentric mass vibrators will be purchased from ANCO Engineers. These systems can be usedindependently, or they can be synchronized, to provide sinusoidal or random loading capabilities and multi-source structural excitation. The two MK-15 shakers, with operational frequencyrange of 0 to 25 Hz, are capable of producing a maximum force of 200,000 pounds (900 kN) when synchronized. The MK-14A unit has a frequency range from 0.1 to 4.2 Hz. In addition, a
linear inertial shaker (i.e., loading produced on a structure with an actuator and a mass that can be mountedon a building) will enable dynamic loading capabilities across a very wide frequency range. Innovativesensors with wireless data transmission capability that allow for rapid installation of high-densityinstrument arrays will be used to provide data streams into the central data recording system (dataacquisition blocks and Antelope software program) purchased from Kinemetrics. Sensors incorporatedinto the system include accelerometers to monitor structural vibrations, accelerometers to monitor soilvibrations during structural tests, linear variable differential transducers and fiber optic sensors for localdisplacement measurements. The data acquisition blocks (each composed of a digitizer, wireless router,power supply and antenna) will transmit the digitized data to a command center through a wireless IPnetwork surrounding the tested structure, where it will be acquired and processed by the Antelope dataacquisition software package installed on a UNIX workstation. These test data can be observed on-lineor off-line with local personal computers networked to the UNIX server. The UNIX workstation will alsostream the data in real-time from the site command center to UCLA via internet satellite communication.Lastly, a Hogentogler cone penetration truck has been purchased to allow for rapid and detailed sub-surface characterization and for the installation of subsurface accelerometers that would be developedas part of the project. The rig is equipped with a seismic-piezocone to permit measurement of soilconsistency and strength, shear wave velocity (through downhole testing), and pore water pressures.
UCLA Mobile Field Laboratory Schematic
Field Testing and Monitoring of Structural Performance
University of California, Los AngelesJohn Wallace, Principal Investigator
Website: http://www.cee.ucla.edu/nees
CONSORTIUM of UNIVERSITIES for RESEARCH in EARTHQUAKE ENGINEERING
CUREE 2002
15
The NEES Equipment Award at the University of Texas involves development of large-scale field equipment aimed at advancing the state-of-the-art in in-situ dynamic material property characterization and field testing of soil deposits and soil-structure systems. The next-generation field equipment includes a large triaxial mobile shaker called a vibroseis, two cubical shakers, field instrumentation, and teleparticipationequipment.
The triaxial vibroseis, manufactured by Industrial Vehicles International, con-sists of an electro-hydraulic shaker that can generate forces in the X, Y, or Zdirections. It is capable of generating a theoretical peak vertical force of 250 kNover a frequency range of 11 to 200 Hz and a theoretical peak horizontal force of125 kN over the frequency range of 5 to 200 Hz. The vibroseis can be used toactively excite the ground surface, foundation elements over which it can bepositioned, and bridges or other structural systems upon which it can be driven.The vibroseis can also be used to passively excite, through ground-bornevibrations, soil deposits, foundations, buildings, and bridges.
The two cubical shakers are stand-alone, electro-hydraulic shakers that are capable of operating in any direction. One shakerwill be used to generate forces of about 18 kN over frequencies from 2.6 to 800 Hz. The other shaker will be used to generateforces of 50 kN over frequencies from 0.7 to 125 Hz and will be capable of generating a force of 4.5 kN at 0.25 Hz. The cubicalshakers are designed to compliment and extend the shaking capabilities provided by the vibroseis. The smaller cubical shakercan operate at very high frequencies, which is required for geophysical reflection and refraction profiling. The larger cubicalshaker offers larger forces in the frequency range of 0.25 to 2 Hz, a critical frequency range in deep surface-wave testing, in situnonlinear soil property characterization, and in situ liquefaction testing.
A field instrumentation van will house the control systems and instrumentation for the mobile field equipment. The van will carrydata acquisition systems, waveform processing equipment, computer workstations, sensors, and teleparticipation equipment.Sensors will include 3-D 1-Hz geophones, additional smaller geophones, accelerometers, and pore pressure transducers. Bothwired and wireless data acquisition systems will be employed, along with telepresence capabilities. Telepresence will allowremote researchers to interact with field personnel and view the results from field testing.
Natural geotechnical materials represent the largest fraction of all materials that impact the performance of the built environmentduring earthquakes, but these materials are the least investigated, most variable, and least controlled of all materials in the builtenvironment. The NEES equipment under development at the University of Texas will improve our ability to fully characterizethe subsurface and will significantly advance our fundamental understanding of earthquake effects associated with geomaterials.
Triaxial vibroseis from IndustrialVehicles International
Large-Scale Mobile Shakers and Associated Instrumentation forDynamic Field Studies of Geotechnical and Structural Systems
University of Texas at AustinKenneth Stokoe II, Principal Investigator
Website: http://www.geo.utexas.edu/nees
Schematic of field-testing setup and results from in situ liquefactiontesting and in situ nonlinear soil property characterization.
