Cold-formed steel construction performs better than expected in seismic tests
July 25, 2014
ANCHORAGE, Alaska - Engineering researchers have provided the building blocks necessary for enabling performance-based design for cold-formed steel buildings, structures that have shown in shake-test experiments at the State University of New York at Buffalo to withstand seismic loading much better than previously expected.
Light, strong and easy to construct cold-formed steel (CFS) buildings are repetitively framed with light steel members and conform to well-defined seismic design codes.
Until this study, however, engineers and builders significantly underestimated the seismic strength of cold-formed steel structures. In fact, following the shake-table experiments at maximum-considered earthquake levels, little to no damage to the structural system was observed and the test specimen had no residual drift.
"There is a large difference between the idealized engineering models of the seismic lateral force resisting system and the superior performance of the full CFS building system," said project leader Benjamin Schafer, professor and chair of the Department of Civil Engineering at Johns Hopkins University. "In other words, we've shown that CFS structures hold up extremely well under earthquake conditions and that it is possible to design CFS structures even more efficiently."
Funded by the George E. Brown Jr. Network for Earthquake Engineering Simulation (NEES) and the steel industry, the project is aptly named CFS-NEES. The CFS-NEES findings were discussed in two papers presented during Quake Summit 2014, which is part of the 10th U.S. National Conference on Earthquake Engineering on July 21-25 in Anchorage.
The study produced basic data on the hysteretic performance of connections, members, assemblages and full CFS buildings, enabling true "performance-based design" for CFS buildings. Performance-based design (PBD) is an efficient and increasingly important technique that allows engineers to design structures to withstand specific seismic loads.
"Our aim in the CFS-NEES project has been to develop experimental benchmarks, fundamental characterizations and generally demonstrate efficient means for modeling cold-formed steel structures, which are inherently complex," Schafer said.
Until this study, current seismic response systems primarily used mechanisms independent from CFS members, such as bearings or straps, to resist seismic demands.
To acquire the key data on CFS member response to earthquake shaking stress (commonly referred to as cyclic loading), the American Iron and Steel Institute in collaboration with the CFS-NEES effort, funded a companion project to provide explicit data on the cyclic response of cold-formed steel members.
The results of these tests, conducted at Virginia Polytechnic and State University, enabled the team to develop seismic force resisting systems that incorporate complete cold-formed steel member-response. These systems were included in the study's full-scale experiments.
Another collaborative component of the work took place at the University of North Texas, where the team tested the strength of Oriented Strand Board sheathed shear walls. OSB is a type of engineered wood particleboard commonly used for load-bearing applications, including for CFS structures.
Those tests revealed that in nearly all cases, developed strength of the shear walls was in excess of current code predictions. Furthermore, the addition of other construction elements, such as interior gypsum, can lead to even greater than expected "overstrength" in shear walls.
In summer 2013, the project team conducted tests of two full-scale, two-story CFS-framed buildings using the three-directional twin shake tables at the University at Buffalo NEES laboratory.
"The CFS-NEES project is the first to test a full CFS building designed to North American specifications," said Kara Peterman, a postdoctoral research associate at Northeastern University. "The building specimens were designed as a CFS archetype - intended to be representative of modern cold-formed steel practices for commercial construction."
The buildings were nominally structurally identical with several key differences. The initial, or Phase 1 structure, did not include nonstructural components and was framed with just the lateral force-resisting system and the gravity system. The second structure, Phase 2, was ultimately framed with nonstructural elements including exterior OSB, interior drywall, stairways, interior partition walls and exterior weatherproofing (DensGlass).
The buildings were densely instrumented with cameras and sensors that provided video, displacement, acceleration and force measurements both globally and in local systems throughout. The general aim of the sensors was to capture the building motion, multistory shear wall behavior, floor diaphragm motion and behavior, building system identification, load transfer mechanisms to and amongst shear walls, and participation of the gravity and nonstructural systems.
The Phase 1 testing of the structural system exceeded predictions from prior shear wall tests and OpenSees modeling efforts. The Phase 2 building specimens likewise exceeded performance predictions and design minimums. Notably, the addition of exterior OSB sheathing had the most significant effect on overall performance. Ultimately, during deconstruction of the final specimen, the team detected no damage to the structural system except for minor bubbling of the CFS strap at panel seams on the interior face of the shear walls.
Somewhat to the research team's surprise, the CFS building is stiffer and stronger than current engineering designs suggest; the building responds as a system, not as a set of uncoupled shear walls and the gravity system contributes to the lateral response.
Full-scale testing of the CFS-NEES building provides a first look at the full system effect for buildings framed from cold-formed steel, and the system effect is significant across the board, requiring new approaches in prediction and design. Work remains to address details not fully explored - semi-rigid diaphragm behavior, for example - to fully enable engineers working in this domain.
Quake Summit is the annual meeting for the National Science Foundation's George E. Brown Jr. Network for Earthquake Engineering Simulation (NEES), a shared network of laboratories based at Purdue University. Schafer and Peterman each presented Quake Summit papers, which were written jointly by the team.
The CFS-NEES Effort: Advancing Cold-Formed Steel Earthquake Engineering
B.W. Schafer1, D. Ayhan2, J. Leng3, P. Liu2, D. Padilla-Llano4, K.D. Peterman13, M. Stehman3, S.G. Buonopane5, M. Eatherton6, R. Madsen7, B. Manley8, C.D. Moen9, N. Nakata10, C. Rogers11, C. Yu12
1Professor, Johns Hopkins University; 2Visiting Student Scholar, Johns Hopkins University; 3Graduate Research Assistant, Johns Hopkins University; 4Graduate Research Assistant, Virginia Polytechnic and State University; 5Professor, Bucknell University; 6Assistant Professor, Virginia Polytechnic and State University; 7Senior Project Engineer, Devco Engineering; 8Regional Director, American Iron and Steel Institute; 9Associate Professor, Virginia Polytechnic and State University; 10Assistant Professor, Johns Hopkins University; 11Professor, McGill University; 12Associate Professor, University of North Texas; 13Postdoctoral Research Associate at Northeastern University
The objective of this paper is to summarize a multi-year effort to advance our understanding in the seismic behavior of, and improve the design of, buildings framed from cold-formed steel (CFS). The effort includes a U.S. National Science Foundation funded project and companion industry-funded projects taken together under the abbreviated name: CFS-NEES. Major deliverables in the CFS-NEES effort include: experimental shear wall testing, characterization, and modeling; experimental cyclic member testing, characterization, modeling, and design; and, complete building design, modeling, and shake table testing. The research enables performance-based design by providing the necessary building blocks for developing nonlinear time history models of buildings framed from cold-formed steel. In addition, the experiments demonstrate the large difference between idealized engineering models of the seismic lateral force resisting system and the superior performance of the full building system. Significant work remains to bring the findings to design practice, and this effort is both ongoing and an area of future need.
Seismic Performance of Full-Scale Cold-Formed Steel Buildings
K.D. Peterman1, M. J. J. Stehman6, S. G. Buonopane2, N. Nakata3, R. L. Madsen4, B. W. Schafer5
1 Postdoctoral Research Associate, Northeastern University; 2Professor, Department of Civil and Environmental Engineering, Bucknell University; 3Assistant Professor, Dept. of Civil Engineering, Johns Hopkins University; 4Senior Project Engineer, Devco Engineering; 6Graduate Research Assistant, Johns Hopkins University; 5Professor and Chair, Dept. of Civil Engineering, Johns Hopkins University
The NSF NEESR project: Enabling the Performance-Based Design of Multi-Story Cold-Formed Steel (CFS) Structures, known simply as CFS-NEES, has entered its final year of research. Testing of two full-scale cold-formed steel framed buildings under seismic excitation at the University at Buffalo Structural Engineering Earthquake Simulation Lab (SEESL) was performed in the summer of 2013. The two-story buildings, approximately 7 m x 15.3 m in plan and 5.8 m in height, were tested in two different configurations. In the first, the engineered lateral force resisting system (LFRS), consisting of OSB sheathed shear walls, and OSB sheathed floors/diaphragms was tested—gravity walls were left unsheathed, and interior gypsum on the shear walls and interior walls were absent. In effect, this first configuration examines the LFRS that is specifically designed by the engineer. In the second building configuration the building was completely fit-out, thus the influence of the sheathed gravity walls, interior walls, etc. were all captured, providing insight on the engineered LFRS and the full building system response. System identification tests and earthquake excitations utilizing the Canoga Park and Rinaldi records were both performed. The buildings were densely instrumented and provide video, displacement, acceleration, and force measurements both globally and in local systems throughout. While the response of the entire structure is investigated, the performance of several sub-systems is also of interest, including: the ledger-framing system, floor diaphragm, multistory shear walls, stud-sheathing-fastener connections, and non-structural elements. Aligned with the overall CFS-NEES effort, these experiments will also provide benchmarks for advancing the computational models necessary for improving performance-based design for CFS structures.