Shaking up ceiling systems will lead to safer structures, more resilient cities
NEESR-GC: Simulation of the Seismic Performance of Nonstructural Systems, Award #0721399
Principal investigator: E. "Manos" Maragakis, Dean of Engineering & Professor, University of Nevada, Reno
If you’re in an office right now, take a look above your head. You’ll see ceiling grids, tiles, lights, and sprinklers – multiple nonstructural components. They may look stable, but an earthquake easily could shake and destroy these fixtures. Even low-intensity earthquakes, those that leave a building’s structure intact, can still wreak havoc on these delicate fittings.
In other words, a building’s nonstructural components -- the interior parts that do not provide support the way building beams, walls, and other structural elements do -- are particularly vulnerable during an earthquake.
Earthquake engineering research led by Manos Maragakis, dean and professor of engineering at the University of Nevada, Reno, has recently completed a major examination of the seismic responses of nonstructural systems. His project goal is to provide engineers and architects with reliable tools and guidelines for improving designs of nonstructural systems.
Specifically, Maragakis and his team focused on ceiling-piping-partition systems: identifying their major failure mechanisms, quantifying how their failure affects critical facilities and metropolitan areas, and producing reliable simulation tools for predicting nonstructural system behaviors in future earthquakes. Ultimately the work will enable engineers to improve designs for fixtures and fasteners, as well as installation methods and construction techniques.
In the United States alone, nonstructural systems represent over 75% of the loss exposure for buildings due to earthquakes, and nonstructural systems account for more than half of the total estimated national annualized earthquake loss.
Even so, engineering and construction communities do not understand the exact behaviors of nonstructural components during earthquakes. No wonder things like fire sprinkler piping, wall partitions, and hanging ceilings continue to fail during earthquakes, leading to expensive building damage, non-functional structures, and of course, injuries due to falling debris.
Noted Maragakis, “Earthquakes always catch us by surprise. We see large-scale nonstructural damage and we wonder exactly how it was originated and propagated. When studying earthquakes, we’re continually learning.”
To address the problem of nonstructural system failures, Maragakis and his multi-disciplinary team conducted a systematic, six-year study of the behavior of ceiling-piping-partition systems during seismic events.
It is the first project to adopt an integrated approach, one that includes both subsystem and system-level experiments. Not only did the team test pipe, ceiling, and partition subsystems, they tested these elements all together to understand how they interact with each other, and at the same time, how they interact with their supporting structure.
These tests will enable the team to develop more reliable analytical models and other simulation tools of nonstructural components and systems. Validated computer models will not only explain and predict the seismic behavior of nonstructural systems, they can be used to design practical methods for earthquake-proofing ceiling-piping-partition systems in modern buildings.
This effort was funded by the George E. Brown Jr., Network for Earthquake Engineering Simulation (NEES). NEES is an NSF-sponsored network of 14 simulation equipment sites and a centralized cyberinfrastructure for sharing data and simulation tools.
Primary objectives. Using specialized earthquake simulation equipment at NEES facilities, Maragakis and his team performed shake tests on existing and newly developed designs for ceiling, piping and partition components, identified failure points, and observed how these items interacted with the structures they were installed in.
Modeling behaviors. In tandem with their physical experiments, the research team developed analytical models for ceiling, partition, and piping subsystems. For example, the piping models incorporate a variety of commonly used components such as distribution mains of various pipe diameters, branch lines of various pipe diameters, hangers, seismic braces, wire restraints, sprinkler heads, and more than 900 nonlinear threaded and grooved joint springs.
Using SAP software, the team developed a refined 3D model for ceiling systems to perform a comparative set of fragility analyses for differing geometries and bracing conditions of suspended ceiling systems with square lay-in acoustic tiles.
Also, prior to building their test structures, the team used the “OpenSees” simulation, modeling and analysis software to develop a full 3D model of the test-bed structure.
Shaking components and subsystems. This research project required the team to construct full-scale sample specimens and shake them on specialized, large-scale earthquake simulation equipment. Literally, earthquake researchers shake their test specimens to failure so they can understand exactly how components and systems fail and break.
The experiments took place at multiple NEES network sites, including the University at Buffalo and at the University of Nevada, Reno.
At Buffalo, the Nonstructural Component Simulator (NCS) was used to simulate full-scale horizontal floor motions on selected piping and partition subsystems. In an earthquake, buildings tend to amplify the shaking from the ground to the roof, and the NCS can reproduce the larger motions occurring in the upper stories.
Also, researchers evaluated the behavior of full-scale sprinkler piping tee joints to discover how much the pipes can deform until leakage occurs or they break completely. Also at Buffalo, two 50-ton capacity shake tables were used for simulating horizontal and vertical excitations on several large lay-in ceiling subsystems. They tested three different ceiling systems, one 16-feet square, one 20-feet square, and one hefty, 20 by 50 foot ceiling.
The earthquake simulations at Buffalo revealed several types of subsystem failure, including fractures in CPVC cemented joints and the vertical hanger pull-outs in piping systems, as well as connection failures in ceiling systems.
Shaking up the system. The final phase of the project was a series of full-scale shake table experiments at the University of Nevada, Reno (UNR). The idea was to shake a complete system, an entire, large-scale structure filled with nonstructural components, very much like a real building.
For the test-bed specimen, the research team devised an innovative, reusable building. The two-story, two-bay-by-one-bay structure was 60 feet long, 24.5 feet high and 11.5 feet wide. It was outfitted with 560 square feet of suspended ceiling with lay-in tiles, 190 feet of partition walls, and a sprinkler piping system with several branch lines on each floor.
All these nonstructural components were densely instrumented with a variety of over 400 sensors, including accelerometers and displacement and force-measuring devices. Data collection systems and broadcasting equipment were also attached to ensure proper collection and storage of the test data.
The structure was set on 3 of UNR’s 14-square-foot, 50-ton-capacity shake tables. These computer-controlled, hydraulically driven tables simulate earthquakes with as much as twelve inches of displacement side-to-side and shake with a maximum velocity of 40 inches per second.
Broad impacts on building design
The findings and data generated by these experiments will provide a comprehensive understanding of the seismic behavior of ceiling-piping-partition nonstructural systems. They also will result in state-of-the-art simulation tools. Furthermore, the study will provide compelling public policy arguments for amending building codes and standards. Data from the study, for example, prove that the cost of earthquake losses due to nonstructural damage is much larger than the cost of retrofitting existing nonstructural systems or implementing new design methods to reduce the damage.
The project’s robust fragility database for components and sub-systems of the ceiling-piping-partition system will soon be available for practitioners to use on the NEES cyberinfrastructure platform; and the findings will be incorporated into a performance-based framework such as ATC-58, a national rubric for developing modern, performance-tested seismic design procedures for new and existing buildings.
Maragakis anticipates further exploration of nonstructural systems. “These experiments have guided us well. Currently we are preparing specific recommendations for continuing research,” he said.