BGS Group Poster

1
Bolted Semi-rigid Connections
Damage of Welded Connections During Northridge
Earthquake (1994)
As an alternative to moment frames with fully
rigid connections, the frames can be designed
using semi-rigid connections. This type of
system, shown in the figures below, has not been
used extensively in earthquake design mainly
because of its relatively large flexibility,
which could lead to larger global deformations
when compared to fully-rigid frames. However,
experiments have shown that semi-rigid
connections can provided adequate drift control,
reasonable ductility capacity and can dissipate
energy in a stable way. Moreover, semi-rigid
connections are advantageous as they do not use
welds which are prone to crack initiation. Other
advantages of this type of connection compared to
welded connections are its lower cost, and
simplicity of fabrication and field installation.
Further, this type of connection can be designed
as the weak point of the frame, implying that the
column no longer needs to be designed such that
its moment capacity exceeds that of the beam as
in the strong-column weak-beam philosophy. This
can result in more economical structures.
During the Northridge earthquake of 1994 many
steel buildings suffered significant damage.
This damage was especially prevalent in the
welded rigid beam-to-column connections of the
steel moment frames. Most of these failures were
due to brittle fracture initiated at the welds
and did not behave in the ductile way originally
intended. The buildings did not collapse but the
economic losses associated with the direct and
indirect costs due to repair were very
significant.
Conceptual Bending Moment / Joint Rotation Curves
for Various Connections
Fracture of Beam-to-Column Connections
Extended-End-Plate
Double Flange-Angle
T-Stub
Substructured Pseudodynamic Simulation
NEES Multi-Axial Full-Scale Sub-Structures
Testing and Simulation (MUST-SIM) Facility
The magnitude of the seismic loads which a
structure is subjected to depends not only on the
earthquake characteristics but also on the
properties of the structure itself such as the
fundamental period of vibration. As such, it is
important to test the whole structure subjected
to an earthquake instead of applying
quasi-statically a predetermined displacement
time history to the specimen. The use of
pseudo-dynamic substructure testing allows
experimental testing of the connections while
simultaneously considering the effect of the
system behavior under realistic seismic loads
through a computational model of the entire
structure. This experiment will be conducted at
the UIUC MUST-SIM facility.

• The primary objective of MUST-SIM is to create a
  facility in which a full-scale subassembly can be
  subjected to complex loading and imposed
  deformation states at multiple connection points
  on the subassembly, including the connection
  between the structure and its foundation.  The
  MUST-SIM facility will have the following unique
  features
• 6-DOF load and position control at multiple
  connection points
• System modularity to allow for easy expansion and
  low-cost maintenance/operation
• Multiple dense arrays of non-contact measurement
  devices
• Advanced visualization and data mining
  capabilities for integrated teleoperation and
  teleobservation

• Full Scale Facility
• 50-0x30-0x28-0 Strong Wall
• 2-0 Strong Floor (Box Girder)
• 40ton Crane
• Telepresence Telecontrol

Measured Forces (F)

• 1/5th Scale Facility Preliminary Testing,
  Demonstration Training
• Aluminum model strong wall and strong floor
• (3) Model LBCBs
• (3) 6 Model Load Cells
• Telepresence Telecontrol

Simulation Coordinator (UI Sim-Cor)
Target Displacements (u)
(3) LBCB Load Boundary Condition Boxes

• (3) Load Cells
• Capable of measuring all 6 degrees of freedom
• Mounted between LBCB and specimens
• Forces fed back into UI Sim-Cor

Computational Component
Experimental Component