Cyclic performance of haunch repaired steel moment connections: experimental testing and analytical modeling

摘要

The use of a haunch on the bottom side of the beam to enhance the seismic performance of damaged steel moment frame connec- tions has been proposed after the 1994 Northridge earthquake. The effectiveness of such a scheme was demonstrated through both static and dynamic testing of four full-scale specimens. Four full- scale pre-Northridge moment connection specimens were tested either statically or dynamically to failure. It was found that the per- formance of dynamically loaded specimens tended to be inferior to that of the statically loaded specimens. Furthermore, the fracture pattern was also different. Depending on the damage pattern, these specimens were then repaired with a haunch and retested. Static and dynamic test results showed that the cyclic performance could be significantly improved, especially when some modifications were made to the existing groove welded joint of the top flange, and the plastic hinging of the beam occurred away from the face of the column. The presence of a haunch complicates the behavior of an enlarged (or ‘dual') panel zone. To predict the inelastic seis- mic response of a steel framed structure, an analytical procedure for the modeling of a dual panel zone is presented, the simplified analytical modeling is shown to correlate well with the test results.