After the 1994 Northridge earthquake, many girder-to-column connections were found to have fractured. The reasons for this kind of failure are not well understood among the engineering community. Consequently, there has since been a tendency to over-specify the connection details. However, research has shown that this might not insure better connection behavior, while it does increase fabrication cost.
Deformed shape of a cruciform specimen
Matt Christensen, Undergraduate Student Researcher
Sean Cotton, Graduate Student Researcher
Paul Pilarski, Graduate Student Researcher
Sara Prochnow, Graduate Student Researcher
Yanqun Ye, Graduate Student Researcher
This project is studying the seismic and non-seismic connection behavior and developing recommendations for design of these connections. This work is currently focusing on the design of column transverse stiffeners and web doubler plates. The primary components of this project include a literature review on the design criteria and previous research work, experimental research, and parametric finite element analyses to corroborate the experimental work and predict the performance of various connection details.
Analysis work is being done with abaqus. The researchers are conducting three-dimensional nonlinear finite element analyses to study the different connection details. These analyses are modeling all specimens tested in the experiments as well as conducting appropriate parametric studies. The model geometry, material properties, boundary conditions, and loading history reflect the true experimental configurations. The results from the analyses, such as the strain distributions, strain range, and stress concentrations, are used to corroborate experimental work. Extended parametric studies results are then used to help predict connection behavior that is not covered in the tests.
These researchers have been initiating the parametric studies by exploring the various features provided by the analysis program. Both the advantages and limitations for the program have been carefully studied. This helps the researchers in choosing the correct model and the analysis procedure that can best represent the actual experimental work.
Modeling of the pull plate experimental test specimens has already been done. By using symmetry, only one-quarter of an actual specimen was modeled. Three-dimensional, eight noded solid elements were then used to model all the connection parts. Static analysis was used, including geometry and material nonlinearity. Large deflection and small strains were assumed. A displacement loading history is now being applied to the pull plate ends according to that used in the tests. Monotonic loading histories are being modeled through the use of load steps and increments.
|
|
URL: http://www.msi.umn.edu/about/publications/annualreport/ar2000/depts/IT/CivEng/dexter.html |
|
| This page last modified on Friday, 30-May-2008 16:14:08 CDT | ||
| Please direct questions or problems to help@msi.umn.edu | ||
|
Website related questions or problems should be directed to
webmaster@msi.umn.edu
The Supercomputing Institute does not collect personal information on visitors to our website. For the University of Minnesota policy, see www.privacy.umn.edu. © 2001 by the Regents of the University of Minnesota |
||