Objective - To establish a new generation of performance-based design assessment criteria that captures a component action’s dependence on loading history by conducting a set of focused experimental tests and computer simulations on eccentrically braced frame link beams.
What is the technical idea? Performance-based seismic design has gained traction in the U.S. building industry as an alternative way to design new buildings to resist seismic effects. The current standardized performance-based design methodology is contained within ASCE/SEI 41 – Seismic Evaluation and Retrofit for Existing Buildings (ASCE 2013). ASCE/SEI 41 takes a component level approach in assessing the seismic performance. The challenge is that current permissible deformation limits used in the assessment criteria are largely based on fully-reversed cyclic tests, which is often the only loading protocol explored given the cost of experimental testing. The permissible deformation limits do not explicitly account for the influence of a component capacity’s dependence on deformation history, even though it is well-known that a typical component loaded monotonically will have a much different capacity than a component loaded cyclically (Krawinkler 1983, Krawinkler 1996). Recent research has highlighted that some component level actions do not normally experience fully-reversed cyclic deformation demands, especially when rare, collapse-level events are investigated (Maison 2016). This suggests that a new paradigm needs to be established for performance-based assessment procedures in ASCE/SEI 41. It is proposed that components should not be assessed for maximum values, but rather cumulative-based values which directly capture deformation history. To complement this component-level cumulative approach, new limits should also be established for acceptable global system performance to ensure reasonable behavior.
To achieve these goals, an experimental testing program will be conducted on eccentrically braced frame link beams as describe in the next section. An eccentrically braced frame is a lateral force-resisting system in which the brace centerlines do not intersect at the beam level, therefore a shear “link beam” is formed and is expected to be a structural “fuse” under strong ground shaking. The advantage of such a system is it provides the stiffness typical of a braced frame and the ductility typical of a moment frame. Moreover, damage is intended to be focused in the fuse which can be replaced following strong shaking. Researchers have investigated the effects of loading protocols on link beams (Okazaki 2007, Richards 2006), but their focus was mainly on design-level rather than collapse-level earthquakes, which resulted in mostly fully-reversed cyclic response being investigated.
What is the research plan? This research will rationally address the above technical challenges by undertaking the following tasks: (a) conduct a series of experimental tests on eccentrically braced frame link beams using multiple loading protocols, (b) conduct complementary analytical simulations using nonlinear finite element software to extend the results of the testing scheme to configurations not tested, (c) using the results gleaned from (a), (b), and existing literature, formulate an improved set of assessment criteria for eccentrically braced frame link beams that explicitly captures loading history.
Regarding Task (a), a test-setup will be designed and fabricated to facilitate flexibility in specimen size (both length and cross-section). A total of 24 tests (4 specimen sizes, 6 tests each) are initially anticipated to be performed in the structural testing laboratory in EL. Each specimen will be subjected to a set of loading protocols, including fully-reversed cyclic (often considered the “default” testing approach), one-sided cyclic, and monotonic protocols. One protocol will be repeated three times for the initial specimen size to investigate the variation in behavior. Repeat tests for the remaining specimen sizes will be conducted based on the findings. Low cycle fatigue behavior will be investigated by using the protocol suggested in FEMA 461 (FEMA 2007). Pending availability of laboratory space and equipment, the test setup may also be able to apply axial load, which has been shown to affect the rotation capacity of a link beam (the testing program will be expanded if this capability is realized).
This research will have an auxiliary goal of addressing the uncertainty of full-scale experimental measurements. The experimental tests will be devised in such a way to produce redundant measurements and repeated tests (as described above) of identical specimens to further glean uncertainty information. The steel mill certificates will be reviewed to quantify the range and uncertainty of material properties. This information with be fed into a concurrent project studying the propagation of uncertainty in earthquake engineering analysis (project leader: Sattar).
To extend the experimental results from Task (a), Task (b) will involve constructing nonlinear finite element models to simulate the response of the link beams. The models will provide a valuable pre-test prediction of the response and then will be calibrated to match the actual test results. The results will be compared to other test results available in the literature and then used to establish trends of link beam capacity by extending the experimental results obtained in this project. Furthermore, collaboration with Dr. Hussam Mahmoud of Colorado State, a member of the NIST CoE, will be undertaken to expand the impact of this work. Dr. Mahmoud has developed analytical results for buildings with eccentrically braced frames using link beam properties currently found in the literature. The results of this work will be employed by Dr. Mahmoud to ascertain the impact of this new link beam data and acceptance criteria using his analytical approach.
The planned outcome of Tasks (a) and (b) is an improved set of acceptance criteria for link beams, which will be formalized in Task (c). The criteria will directly account for loading history by using the components cumulative plastic deformation capacity as an input parameter. By doing this, the new criteria will provide a more fundamentally-sound basis for assessing a components performance while still maintaining simplicity of the current ASCE/SEI 41 approach. Further, a secondary outcome of this project will be a new framework for deriving experimentally validated cumulative-based assessment criteria. This framework will have the potential of being applied to additional components and systems.