Nonlinear Investigation - Nonlinear Model
blueline.JPG (929 bytes)
The Ji Lu Elastic Models covered variations in boundary conditions and member properties.  For all the models listed, the connection of the main pylon was modeled with complete fixity.  The end pier columns to foundation connections were also modeled with total fixity.  
The cables were modeled as a single line element connecting the tower with the deck.  There mass was lumped by the program at the two connecting nodes for each cable element.  In all cases the pylon was modeled with cracked properties found from section analyses below and above the deck.  In the structure, the pylon above deck tapers continuously from 6 meters in the longitudinal deck direction at the base to 4 meters at the top of the bridge.  This was modeled as 6 meter from the deck surface to half way up the pylon.  Then 4.6 meters the rest of the way.  The behavior of the pylon in the transverse direction is not critical to the behavior of the overall system so this approximation is justified.  The pylon above the deck was modeled with crack properties for the elastic model but in the nonlinear analysis, effects of an uncracked assumption was investigated.
Figure 1 - SAP2000 Elastic Model

 

 

 

Forces in the cables were checked against capacities and there was no concern of cable fracture.  There was, however, a cable fracture found in the Ji Lu after the earthquake but its causes have already been established.  The cables were considered tension and compression elements for they were assumed to have stiffness thought the duration of the earthquake.  The prestressing force in the cables at the time of the earthquake was sufficient to ensure the cable never realized a compression force and thus, maintained their tension and compression stiffness.  The effect of the cable drape on the tension stiffness which is important in modeling correctly elastic dynamics and initial states of the bridge was ignored.  This can be justified by the large inelastic displacements that are being sought.  The elastic displacements are small compared to the inelastic.  The final deck level was input into the elastic model as transcribed from the construction drawings.

The changes in boundary conditions at the end supports and variations in the deck bending stiffnesses were considered in the following models.

Model L Free longitudinally at the end spans.  The initial state of the bridge.  Cracked section properties for the deck.
Model LT Free longitudinally and transversely at the end spans representing the behavior after the bridge was dislocated from its lateral shear restraint.  Cracked section properties for the deck.
Model G Connections to the end spans were modeled with gap elements (Figure 2).  Freedom in the longitudinally and upward vertical was allowed.  Freedom in the transverse direction was constrained the elevation of the architectural wall.  Cracked section properties for the deck.  

Figure 2 - End pier modeled with gap elements
Model LTU Free longitudinally and transversely at the end spans.  Uncracked transverse deck bending stiffness.
Model LTUW Free Longitudinally and transversely at the end spans.  Uncracked deck stiffness properties in the transverse direction with the wings acting in compression.


[Elastic Structural Models] [Ground Motion Variation] [Periods of Vibration] [Elastic Study Results]