9416363 Leon Unique failures by fracture to welded moment-resisting connections in steel frames were observed after the Northridge, California earthquake of January 17, 1994. These failures can be categorized either as Ostructural,O which is associated primarily with larger ductilities and local strains in the steel members and welds than is envisioned in design, or Ometallurgical,O which is associated primarily with weld defects, poor weld workmanship, lack of toughness in thick sections of wide- flange members, and the presence of stress risers such as backing plates. The resolution of the primary cause of the failures will determine the provisions of the seismic design specifications which must be modified. It will also dictate the nature of the repair and rehabilitation that will be required in steel frames, since the retrofit strategies will differ depending on the primary cause of failure. This project addresses the problem of structural behavior by examining both the load and resistance aspects of the problem. With respect to resistance, this research will investigate whether the presence of a composite floor slab and the occurrence of yielding in the panel zone region of the connection significantly increased the local strains near the critical welds. A combined analytical-experimental investigation will be conducted to focus on the local behavior of the welded connections, and in particular on the region near the bottom beam flange, where most of the fractures occurred. The experimental program will include tests of three beam-to-column subassemblages, including one with no slab, one with a slab having partially composite action, and one with a slab having fully composite action. Three-dimensional nonlinear continuum finite element analysis will be used to quantify the local strains near the welds and the effect on force transfer of partial composite action at the concrete/steel interface. With respect to loading, this research will address th e probable force distribution in the structure during the earthquake through the use of a three-dimensional, nonlinear, transient dynamic frame analysis program capable of tracking both horizontal and vertical accelerations. This research will have a profound effect on the design of steel structures if it can be demonstrated that current seismic analysis and design methodologies for steel frame lateral-resistance systems are inadequate, since they usually exclude the effect of the composite floor slabs on the structural behavior. This is a Northridge Earthquake project. ***

Project Start
Project End
Budget Start
1994-09-15
Budget End
1996-08-31
Support Year
Fiscal Year
1994
Total Cost
$65,000
Indirect Cost
Name
University of Minnesota Twin Cities
Department
Type
DUNS #
City
Minneapolis
State
MN
Country
United States
Zip Code
55455