SEISMIC BEHAVIOR, ANALYSIS AND DESIGN OF COMPLEX WALL SYSTEMS: PROJECT SUMMARY Reinforced concrete structural walls are used commonly as the primary lateral-load resisting system for new and retrofit construction. However, despite the heavy reliance on wall systems by practicing engineers, recent efforts to develop performance-based design methods have not yet begun to address structural walls. Today, engineers have few resources to consult regarding the simulation of wall response using practical linear and nonlinear numerical models or the prediction of wall damage (e.g., concrete crack width and concrete spalling) as a function of engineering demands (e.g., inter-story drift). In part, the inadequacy of available performance-based design tools for walls is a result of limitations in the available experimental data. Prior to the NEES initiative, experimental facilities were such that only simplified loading conditions, loading histories, boundary conditions, and geometries could be employed in the laboratory. Additionally, instrumentation was such that experimental data typically include wall displacements and end rotations, average strain measurements and a few analog images showing damage progression. Only a few existing experimental data characterize the response of walls with complex geometries or subjected to bi-directional load histories. None of these data sets include high-resolution strain fields for use in validating advanced simulation tools, nor do any of these data sets include highresolution digital images that can be used to quantify damage. No experimental investigations consider boundary conditions that simulate soil and foundation deformations.
INTELLECTUAL MERIT The research proposed here will advance the understanding of, and simulation tools for, the seismic performance of slender walls through experimental and analytical investigations of wall systems with 1) configurations used in modern design, 2) load distributions that are representative of earthquake loading, 3) and consideration of soil-structure-interaction effects. The advanced experimental capabilities of the UIUC MUST-SIM NEES facility will permit realistic simulation of these complex conditions. Instrumentation developed as part of the MUST-SIM facility makes possible high-resolution monitoring of test specimen displacement fields and such data are necessary to enable the proposed model development effort. These data will be used to advance the state-of-the-art for simulation of reinforced concrete structures through the development of fiber-shell elements that can be used to simulate the inelastic response, including localized damage mechanisms, of three-dimensional walls. Results of the proposed experimental investigation and parametric studies using the high-resolution numerical models will be used to advance the state-of-the-practice for design of wall systems including the development of elastic and simplified inelastic modeling techniques that are appropriate for use with commercial software. Additionally, experimental and simulation data will be used to develop performance-based design tools that account for modeling uncertainty and link engineering demand parameters and damage states.
BROADER IMPACT The structural engineering profession recognizes that the limitations identified above represent a significant gap in the development and implementation of performance-based seismic design provisions. To facilitate the transfer of the research result to practice, an External Advisory Panel has been assembled for the project that includes prominent structural engineers and members of structural engineering societies. An interactive website will be developed to disseminate research results to earthquake engineering professionals, educators and students to allow real-time viewing of experimental tests. In addition to educating today's engineers, the research team will use the research process and results to educate future engineers using the tele-observation capabilities at the NEES facility and education modules in design and analysis classes. The project team will leverage its own diversity and established programs in the colleges of engineering to reach a diverse student population. In addition to these activities, the team will work with college of engineering at UW, UICU and MUST-SIM group to reach K-12 students with activities such as Engineering Open House.
Summary of Research Activities The research employed laboratory testing and numerical modeling to advance understanding of the earthquake behavior and design of reinforced concrete (RC) walls and walled buildings. To investigate behavior and develop data to advance simulation and design, eight RC wall specimens were tested using the NEES laboratory at the University of Illinois (https://nees.org). Test specimens were one-third scale and represented the bottom three stories of a ten-story prototype planar, planar-coupled or c-shaped wall. All specimens were tested quasi-statically under axial loading and either unidirectional or bidirectional lateral loading. A large volume of experimental data was collected including load, displacement and deformation data, damage data, and video and still camera images. Figures 1-5 provide details of the experimental test program and collected data. Experimental data show that i) modern slender concrete walls have the potential to exhibit compression-controlled flexural failure, characterized by rapid strength loss, at relatively low drift demands, ii) the potential for compression-controlled flexural failure is exacerbated by increased compression demand resulting from high axial loads (rare), coupling of two walls (common in mid- to high-rise buildings with multiple elevators) or an asymmetric wall cross section (common in low- to mid-rise buildings), iii) slender walls that exhibit significant loss of lateral load carrying capacity typically maintain moderate axial load carrying capacity. Numerical simulation was used also to investigate the earthquake response of RC walls and walled buildings. Experimental data were used to evaluate previously proposed wall models; results showed that existing models did not meet the objectives of accurate, reliable and computationally efficient simulation of response. Experimental data were used to develop a modeling procedure, employing beam-column line elements with fiber-type section models, that met the modeling objectives, including providing accurate simulation of wall response through failure. The model was validated using an extended data set. Figure 6 shows simulated and measured response histories. The validated numerical model was used to investigate the earthquake response of walled buildings and advance seismic design procedures. First, nonlinear analyses of code-compliant walled buildings indicated a likelihood of shear failure, which could be expected to result in rapid strength loss and possibly collapse, at relatively low seismic demand levels. To address this, a capacity-design procedure for shear was developed to ensure adequate shear capacity and minimize the likelihood of shear failure. Second, nonlinear analyses of walls designed using the new shear-design procedure indicated the potential for nonlinear flexural response distributed over the height of the wall. Since, a large volume of transverse reinforcement is required in regions where nonlinear action is expected, it is desirable to isolate nonlinear response to a relatively few known locations. It was found that methods proposed previously by others could be used to address this issue. Third, the FEMA P-695 methodology was used to determine seismic performance factors for use in design of walled buildings; new factors are less than those used currently, indicating that current design methods are unconservative. Finally, numerical simulation was used also to show that foundation flexibility reduces the seismic demands on walled buildings due to the lengthening of the response period of the building and increased damping. Intellectual Merit The research activities advanced experimental methods to develop data characterizing the earthquake response of modern reinforced concrete walls of varying configuration. These data provide unprecedented understanding of behavior and damage progression. Data and supporting documentation are archived in the NEES Project Warehouse (https://nees.org) and available for use by the community. The research activities also advance numerical modeling of concrete walls to enable accurate simulation of response including failure; the proposed modeling techniques employ material and element models implemented in OpenSees, an open source platform for nonlinear analysis (http://opensees.berkeley.edu). Finally, current design procedures for walls were found to be unconservative, resulting in undesirable response modes and an unacceptably high likelihood of collapse under earthquake loading; a new design procedure was developed to achieve desired earthquake performance. Broader Impact The broader impacts of the research grant were extensive. The activities supported by this grant i) contributed to the education of five PhD, six MS, and numerous undergraduate students; all of the graduate students now have positions in academia or industry, and many of the undergraduate students went on to graduate school, ii) advanced the careers and experience of four assistant professors, all of whom received tenure at research universities in the US, iii) broadened participation of under-represented groups in engineering including advancing the university careers of three female assistant professors who are now associate and full professors, one female PhD student who is now a university faculty member, and numerous undergraduate and K-12 students who participated in research and educational programs, iv) disseminated research results to the earthquake engineering community via traditional methods as well as a webinar that was viewed by several hundred people and a YouTube channel (www.youtube.com/user/NEESRWallProject).
