Experimental performance of double built-up T moment connections under cyclic loading

- Full-scale double built-up T connections were tested under cyclic loads.
- Specimens tested exceeded requirements for special moment resisting frames.
- Performance of connection unaffected by type of weld of the built-up T stub.
- Design procedure stipulated by AISC 358 is applicable when using built-up T-stubs.

Prequalification of the Double Split Tee (DST) connection has been conducted using hot-rolled shapes, whose availability is still limited in several countries. In addition, the range of available T web and flange thicknesses is narrow for these shapes. Thus, the use of built-up T-stubs on the DST connection seems to be a promising alternative, considering that they offer more freedom of sizing and allow a better use of the material. The experimental research reported here aimed to validate the use of built-up tees instead of hot rolled tees in DST moment connections. Four beam-to-column full-scale double built-up T (DBT) moment connections were subjected to cyclic loading following the prescribed loading history in AISC seismic provisions. The parameters varied among the specimens were the T flange thickness, the type of weld between the T flange and the stem (fillet weld or full penetration groove weld), and the column web thickness, in order to observe the most significant limit states of the connection, namely development of plastic hinges in beams, panel zone plastification, and T flange prying. The results indicate that the connection can sustain 4% drift without strength degradation, whether it is designed for T flange prying or beam hinging. In the case of the former, only limited permanent deformation of the flange is observed, while for the latter significant plastic deformation of the beam can be achieved with little or no damage of the T. The type of weld used has no significant effect on the connection performance.