Multidisciplinary Center for Earthquake Engineering Research logo google logo
navigation bar

Protective Systems for Buildings: Application of Spherical Sliding Isolation Systems

Abstract
Objective and Approach
Accomplishments
Conclusion
Personnel and Institutions
Collaboration
Publications
Return to Volume Table of Contents


Abstract

Constantinou, M. Protective Systems for Buildings: Application of Spherical Sliding Isolation Systems. Research Accomplishments, 1986-1994: The National Center for Earthquake Engineering Research, pages 55-61. (Buffalo : National Center for Earthquake Engineering Research, September 1994)

The friction pendulum system (FPS) bearing is a form of spherically shaped, articulated sliding bearing. Movement of one part of the bearing with respect to others resembles pendulum motion in the presence of friction. The lateral force needed to induce a lateral displacement, ? consists of a restoring force equal to W?R and a friction force equal to W, where R is the radius of curvature of the spherical sliding surface, ?is the coefficient of friction and W is the vertical load on the bearing. The lateral force is proportional to the vertical load, a property which minimizes adverse torsional motions in structures with asymmetric mass distribution. The stiffness of the bearing (W/R) is proportional to the supported mass so that the period of vibration is only dependent on the radius of curvature and acccleration of gravity. This allows designs with large period of vibration, which has distinct advantages in applications to light weight structures and soft soil conditions.

In 1989, NCEER and Earthquake Protection Systems, Inc. collaborated on the shake table testing of a large, six-story steel moment frame model with FPS isolators installed below a rigid diaphragm. Under moderate to severe level ground motions, no uplift of the bearings was observed despite the large overturning aspect ratio of the model. The isolated structure could sustain, while elastic, a peak ground acceleration about six times larger than it could sustain under non-isolated, fixed-base conditions. The tests also revealed that the isolated structure did not respond as a"rigid block", but rather it exhibited response with higher mode participation, as a result of the nonlinearity of the isolation system and flexibility of the superstructure (nearly one second period in prototype scale).

In 1991, tests were conducted with a seven-story, 47.5 kip steel model under fixed-base and isolated conditions, and with various braced and moment frame configurations. Moreover, tests were conducted with the isolators placed directly at the base of individual columns, rather than having a rigid base above the isolators. In severe seismic loading, large overturning moments developed, which resulted in up to +-100% variation on the axial bearing load. The tests provided a wealth of data, which were used to refine analytical models for the isolation bearings and verify simplified analysis procedures. The analytical model has been later implemented in the computer program 3D-BASIS-ME.

The research on the seismic isolation of buildings, together with research conducted in parallel on the seismic isolation of bridges, established the FPS system as a highly researched, well understood and effective seismic isolation system. As a result of this research, the FPS system was selected for two major seismic isolation projects: the U.S. Court of Appeals building in San Francisco and two liquefied natural gas storage tanks in Greece.

Back to table of contents for this report



The primary objective of this project was to perform an analytical and experimental study of sliding seismic isolation systems. The Friction Pendulum System [FPS] has been given particular attention as being a low profile, compact and highly stable form of isolation system. The research focused on conducting shake table experiments of isolated building models, utilizing a variety of structural and isolation system configurations, and applying severe seismic loading. The experimental results revealed the limits of the effectiveness of the isolation system, and exposed its features, advantages and disadvantages. A second objective was to use the experimental results to develop and verify an analytical model for the bearings, and implement the model in dynamic analysis computer codes.

This research task is part of NGEER's Building Project. Task numbers are 87-2002A, 88-2002a, 89-2101 and 90-2101.

Back to table of contents for this report


Accomplishments

Selected Shake Table Test Results

Testing of the isolated seven-story building model was unique in many ways: for example, isolators were placed directly below the columns without forming a rigid isolation basemat, large overturning aspect ratio, and significant variation in bearing axial load and bearing uplift. Figure 1 shows the friction pendulum bearing. Figure 2 shows a diagram of the moment frame configuration of the isolated model structure. The model is illustrated in Figure 3.


The comparisons show that the isolated structure can withstand earthquake shaking four to six times stronger than the nonisolated structure, while remaining within its elastic drift limit. The bottom story drift is expressed with respect to the exterior and interior column base. This isolation story drift for the isolated structure is much larger than its counterpart, because the column bottom is free to slide and rotate and is not fixed to the shake table as in the fixed-base structure.

The very good performance of the isolated model in the El Centro earthquake was achieved under conditions of severe loading. Figure 4 shows the recorded response of the isolation system in terms of bearing displacement history and loops of shear force vs displacement for individual bearings and for the isolation system. It may be observed that the two exterior bearings (designated as CO and C3) developed shear force at extreme displacement in the range of zero to nearly 2.5 kips. Since the shear force is proportional to the axial load, it indicates a severe variation of the axial load on the exterior bearings, from zero to about twice the gravity load. The CO exterior bearing experienced some limited uplift.

At quarter length scale, the seven-story model with a fixed base (non-isolated) had a fundamental period of 0.45 seconds. When isolated, the bearings had a 9.75 inch radius of curvature, resulting in a 1.0 second rigid body mode period. The coefficient of friction in the bearings under high sliding velocity conditions was 0.06.

One test subjected the model to the 1940 E1 Centro earthquake, component S00E. In the isolated condition, the actual earthquake was scaled up by a factor of 2.0, whereas in the nonisolated condition the earthquake was scaled by a factor of 0.35. In another test, the 1952 Taft earthquake, component N21E, was applied to the two models with scale factors of 3.0 and 0.75, respectively. A comparison of recorded story shear forces and drifts is presented in Table 1.


Story Non-Isolated Isolated Non-Isolated Isolated
El Centro S00E
35%
El Centro S00E
200%
Taft N21E
75%
Taft N21E
300%
Shear
Weight
Drift
Height
(%)
Shear
Weight
Drift
Height
(%)
Shear
Weight
Drift
Height
(%)
Shear
Weight
Drift
Height
(%)
7 0.077 0.231 0.083 0.318 0.06 0.187 0.065 0.194
6 0.138 0.339 0.143 0.42 0.111 0.278 0.097 0.267
5 0.181 0.387 0.159 0.453 0.147 0.34 0.127 0.316
4 0.21 0.458 0.168 0.501 0.184 0.401 0.141 0.358
3 0.218 0.46 0.206 0.569 0.212 0.463 0.148 0.364
2 0.22 0.361 0.226 0.435 0.224 0.36 0.149 0.264
1 0.235 0.281 Ex
0.284 In
0.24 1.369 Ex
0.875 In
0.235 0.328 Ex
0.294 In
0.175 0.753 Ex
0.519 In
Ex: Exterior Column, In: Interior Column, Weight = Total Structural, Weight Height = Story Height
Table 1: Comparison of Response of Non-Isolated Structure


The comparisons show that the isolated structure can withstand earthquake shaking four to six times stronger than the nonisolated structure, while remaining within its elastic drift limit. The bottom story drift is expressed with respect to the exterior and interior column base. This isolation story drift for the isolated structure is much larger than its counterpart, because the column bottom is free to slide and rotate and is not fixed to the shake table as in the ftxed-base structure.

The very good performance of the isolated model in the El Centro earthquake was achieved under conditions of severe loading. Figure 4 shows the recorded response of the isolation system in terms of bearing displacement history and loops of shear force vs displacement for individual bearings and for the isolation system. It may be observed that the two exterior bearings (designated as CO and C3) developed shear force at extreme displacement in the range of zero to nearly 2.5 kips. Since the shear force is proportional to the axial load, it indicates a severe variation of the axial load on the exterior bearings, from zero to about twice the gravity load. The CO exterior bearing experienced some limited uplift.

The U.S. Court of Appeals building in San Francisco

The U.S. Court of Appeals building in San Francisco is an early example of American Renaissance Style and is the only such example in the western United States. Constructed in 1905, it features granite exterior walls, marble columns and statues, handpainted murals and mosaic tile floors. Figure 5 shows the ornate interior and exterior of the building.

The building has plan dimensions of 95 m by 81 m and overall height of 25 m above street level. It consists of a full basement, four floors and two mezzanines. Its weight is approximately 57,000 metric tons. The courthouse survived the great San Francisco earthquake of 1906, but was damaged in the 1989 Loma Prieta earthquake. Damage to the interior tile walls, which formed part of the lateral load resisting system, significantly reduced its seismic resistance. The owner, the General Services Administration, closed the building while investigating retrofitting methods. Seismic isolation was selected over conventional retrofitting schemes because it afforded the greatest degree of seismic protection, had the least impact on the building's historic and architectural character and had the least life-cycle cost.

The FPS system was selected based on cost and technical merit. The technical reasons for the selection of the FPS system have been its extensive testing, its low profile, which allowed installation without cutting away part of the foundation's wood pilings, and its unique construction, which minimized transmission of overturning moment to the footings and averted additional reinforcement of the footings.

The installation of the isolation system was completed in July 1994. It consists of 256 FPS bearings, having a radius of curvature of 1880 mm and displacement capacity of 350 mm. Eight different bearing types carry load in the range of 450 to 5350 kN. Dynamic analyses were performed with the computer program 3D-BASIS.

Liquefied Natural Gas Storage Tanks, Greece

Liquefied Natural Gas (LNG) storage tanks represent critical facilities. Plant-operational and safety aspects dictate designs of these tanks with full containment, as illustrated in figure 6. An inner stainless steel tank contains LNG at cryogenic temperature of about -160C. Approximately 1 m thick insulation surrounds the inner tank. A prestressed concrete outer tank is built around the inner tank and insulation for containing the LNG in the case of rupture of the inner tank, to protect the inner tank against missile or aircraft impact, and to provide support for the insulation and piping systems.

The design of LNG storage tanks is primarily controlled by hydrostatic stresses. LNG has unit weight just less than half that of water. Prior to commissioning, LNG storage tanks are filled with water for testing. This loading condition induces hydrostatic stresses which are larger than the combined hydrostatic stresses caused by LNG fill and hydrodynamic stresses due to moderate seismic excitation. However, above a certain level, seismic loadings begin to become the dominant loading condition. This has been the case for the LNG storage tanks in Greece. Rather than modifying the tank geometry, providing anchorage and accepting yielding of the tank in strong seismic excitation, a situation which may cause inelastic buckling or "elephant foot" buckling, the tanks were designed with an isolation system, which can so reduce the seismic forces that hydrotesting becomes the dominant loading case.

The two LNG storage tanks are currently under construction on the Revithoussa island, which is located near Athens. Both are of the same geometry with an inner tank diameter of 65.7 m and height of 22.5 m. The outer tank has a diameter of 68.7 m and height of 32.3 m. The tanks are positioned in excavated pits 24 m deep and 75 m in diameter. The partial burial is for reasons of aesthetics and is not intended for containment. Figure 6 illustrates the tank construction.

The author is a senior consultant to the owner, responsible for the isolation system and inner tank design and quality assurance of the isolation system. In the preliminary design phase, he developed alternative seismic isolation designs and, together with Professor Andrei Reinhorn and Graduate Student P. Tsopelas, developed the computer program 3D-BASIS-ME for use in the final analysis. In January 1994, the FPS isolation system was selected for the project. The design called for 212 bearings per tank on top of concrete pedestals as illustrated in Figure 6. The bearing design is nearly identical to the largest size FPS bearing of the U.S. Court of Appeals building. Fabrication of the bearings began in April 1994 in San Francisco. Installation is scheduled for early 1995.

Back to table of contents for this report



Conclusion

NCEER and Earthquake Protection Systems cooperated on the experimental and analytical study of the FPS seismic isolation system. The system has been rigorously studied over a period of five years and the results have been promptly disseminated to the engineering community. The system has since been selected for two major seismic isolation projects: the U.S. Court of Appeals building in San Francisco and a pair of liquefied natural gas storage tanks in Greece.

Back to table of contents for this report


Personnel and Institutions

The primary collaborators were Professors Michael Constantinou and Andrei Reinhorn of the University at Buffalo for NCEER and Dr. Victor Zayas for Earthquake Protection Systems, Inc. The work was carried out by University at Buffalo graduate students A. Mokha, T.M. Al-Hussaini and P. Tsopelas.

Back to table of contents for this report



Michael Constantinou
Andrei Reinhorn
University at Buffalo

Victor Zayas
Earthquake Protection Systems, Inc.


Back to table of contents for this report


Publications

Mokha, A., Constantinou, M.C. and Reinhorn, A.M., "Teflon Bearings in Aseismic Base Isolation: Experimental Studies and Mathematical Modeling," Technical Report NCEER-88-0038, National Center for Earthquake Engineering Research, Buffalo, N.Y., 1988.

Mokha, A., Constantinou, M.C. and Reinhorn, A.M., "Sliding Isolated Structures: Experiments and Mathematical Modeling," Proceedings, 1989 ASME PVP Conference, Hawaii, July 1989, Vol. 181, pp. 101-106.

Mokha, A., Constantinou, M.C. and Reinhorn, A.M., "Teflon Bearings in Base Isolation. Part 1:Testing," Journal of Structural Engineering, ASCE, Vol. 116, No. 2, 1990, pp. 438-454.

Mokha, A., Constantinou, M.C. and Reinhorn, A.M., "Teflon Bearings in Base Isolation. Part 2: Modeling," Journal of Structural Engineering, ASCE, Vol. 116, No. 2, 1990, pp. 455-474.

Mokha, A., Constantinou, M.C. and Reinhorn, A.M., "Experimental and Analytical Study of Earthquake Response of a Sliding Isolation System with a Spherical Surface," Technical Report NCEER-90-0020, National Center for Earthquake Engineering Research, Buffalo, N.Y., 1990.

Mokha, A., Constantinou, M.C. and Reinhorn, A.M., "Further Results on the Frictional Properties of Teflon Bearings," Journal of Structural Engineering, ASCE, Vol. 117, No. 2, 1991, pp. 622-626.

Mokha, A., Constantinou, M.C., Reinhorn, A.M., and Zayas, V., "Experimental Study of Friction Pendulum Isolation System", Journal of Structural Engineering, ASCE, Vol. 117, No. 4, 1991, pp. 1203-1219.

Theodossiou, D., and Constantinou, M.C., "Evaluation of SEAOC Design Requirements for Sliding Isolated Structures," Technical Report NCEER-91-0015, National Center for Earthquake Engineering Research, Buffalo, N.Y., 1991.

Tsopelas, P., Nagarajaiah, S., Constantinou, M.C. and Reinhorn, A.M. "3D-BASIS-M: Nonlinear Dynamic Analysis of Multiple Building Base Isolated Structures," Technical Report NCEER-91-0014, National Center for Earthquake Engineering Research, Buffalo, N.Y., 1991

Soong, T.T. and Constantinou, M.C., "Base Isolation and Active Control Technology-Case Studies in the U.S.A.," Proceedings, IDNDR International Symposium on Earthquake Disaster Reduction Technology, Building Research Institute, Ministry of Construction, Tsukuba, Japan, December 1992.

Mokha, A., Constantinou, M.C. and Reinhorn, A.M., "Verification of Friction Model of Teflon Bearings Under Triaxial Load," Journal of Structural Engineering, ASCE, Vol. 119, No. 1, 1993, pp. 240-261.

Constantinou, M.C., Winters, C.W. and Theodossiou, D., "Evaluation of SEAOC/UBC Analysis Procedures. Part 2:Flexible Superstructure," Proceedings, ATC-17-1 Seminar on Seismic Isolation, Passive Energy Dissipation, and Active Control. March 1993, San Francisco, CA.

Winters, C.W. and Constantinou, M., "Evaluation of Static and Response Spectrum Analysis Procedures of SEAOC/UBC for Seismic Isolated Structures," Technical Report NCEER-93-0004, National Center for Earthquake Engineering Research, Buffalo, N.Y., 1993.

Tsopelas, P., Nagarajaiah, S., Constantinou, M.C., and Reinhorn, A.M., "Nonlinear Dynamic Analysis of Multiple Building Base Isolated Suctures," J. Computers and Structures, Vol. 50, No. 1, 1994, pp. 47-57.

Al-Hussaini, T.M., Zayas, V.A. and Constantinou, M.C.,"Seismic Isolation of Multi-Story Frame Structures Using Spherical Sliding Isolation Systems," Technical Report NCEER-94-0007, National Center for Earthquake Engineering Research, Buffalo, N.Y, 1994.

Tsopelas, P.C., Constantinou, M.C., and Reinhorn, A.M. "3D-BASIS-ME: Computer Program for the Nonlinear Analysis of Seismically Isolated Single and Multiple Building Structures and Liquid Storage Tanks," Technical Report NCEER-94-0010, National Center for Earthquake Engineering Research, Buffalo, N.Y., 1994.

Back to table of contents for this report

 


  Contact Us  |  Acknowledgements   |  Disclaimer  |  Copyright© 2007 by the Research Foundation of the State of New York. All rights reserved.