Notes • Comments • Notes • Comments
Thank you to selection committee
Honor to be selected among the fine papers submitted from fellow MCEER students
While steel truss bridge have performed well from preventing collapse, significant damage has left the bridge unusable until repairing could be done.
As part of a past MCEER study funded through the FHWA, investigating the seismic resiliency of steel bridges, some critical deficiencies of steel bridges were identified.  They include….
Recently, a significant experimental testing program was completed at the University at Buffalo investigating the inelastic, cyclic behavior of these built-up members.
This research began with determining the pull-out capacity of the anchorage connections of some existing steel truss bridges.
While gross area yielding of the anchor bolts is a ductile failure mode, yielding will occur on the first cycle initiating uplift but compressive yielding does not occur leaving residual deformations on the anchor bolts
Release of the anchorage connection such that horizontal movement is restrained while allowing uplifting to occur
dissipation such that response parameters deemed critical can be controlled
Unbonded braces were developed in Japan in the 1980’s but have only recently been used as a seismic protective device.  A few manufacturers now exist in the U.S.
A controlled rocking approach to the seismic resistance of bridges was implemented into the design of the South Rangitikei Rail Bridge located in New Zealand and the Lions gate bridge, located in Vancouver British Columbia.  Both used steel yielding devices at the anchorage interface to control the rocking response.
1-2  pre-uplift, “fixed-base” response as the gravitational restoring force is overcome
2-3  pier leg begins to uplift and pull brace until the brace yields at pt. 3. 
        post-uplift stiffness defined by rocking stiffness, kr, which includes flexibility of the tower and rotational flexibility at the base
3-4  unbonded brace yields in tension to the maximum pier displacement at 4.
4-5  pier begins to move in opposite direction, tensile force in the unbonded brace is released and begins to go into compression until it begins                compressive yielding at pt. 5. 5-6  unbonded brace goes through compressive yielding until coming back into contact with the foundation at pt. 6.
6-1  pier deformations are returned to the undeformed position
6-7  After the brace goes through
A key factor for the design of any displacement based hysteretic system is prediction of maximum deformations.
Method 1 & 2 are very similar to the coefficient method of FEMA 356.
2nd cycle properties are used due to the increased flexibility during 2nd cycle response.
Syn. Motions generated using software generated by Engineering Seismology Laboratory at UB
Varying aspect ratios, using a set of pier properties assumed representative, and varying seismic demand
Conservative nature of Method 1 due to use of kr as effective stiffness
Method 2 not bad but unconservative in some cases
Observed fluctuation of base shear around expected static response
Beginning to see effects of rocking, in particular the effects of impacting and uplift from the foundation
Note the use of 1st cycle response for determining maximum forces.  Analysis has shown a decrease in the dynamic effects with the increased flexibility during the 2nd cycle.
In an attempt to quantify the effects, simple linear elastic models are used to represent the application and transfer of these loads.
For uplift, the vertical shearing mode is used.
For impacting of the tower leg, axial deformations of the tower leg is used.
Can think of simple model of dropping mass onto spring.
Notice use of 2nd cycle properties for the spectral capacity to demand procedure.
As excitation ceases, system oscillates about the undeformed position due to self-centering ability.