Multi-Scale Computational Simulation of Progressive Collapse of Steel Frames.

Abstract

Progressive building collapse occurs when failure of a structural component leads to the failure and collapse of surrounding members, possibly promoting additional collapse. Global system collapse will occur if the damaged system is unable to reach a new static equilibrium configuration. The objective of this research is to identify and investigate important issues related to collapse of seismically designed steel building systems using multi-scale computational models.

Coupled multi-scale finite element simulations are first carried out to investigate the collapse response of moment resisting steel frame sub-assemblages. Simulation results suggest that for collapse resistant construction, designers should strive to use a larger number of smaller beam members rather than concentrate resistance in a few larger members and should specify ASTM A-992 steel rather than specifying generic steels. Improved behavior can also be achieved by increasing the shear tab thickness or directly welding the beam web to the column.

Using information gleaned from the sub-assemblage simulations, computationally efficient structural scale models for progressive collapse analysis of seismically designed steel frames systems are developed. The models are calibrated and utilized within the context of the alternate path method to study the collapse resistance of multistory steel moment and braced frame building systems. A new analysis technique termed “pushdown analysis” is proposed and used to investigate collapse modes, failure loads and robustness of seismically designed frames. The collapse and pushdown analyses show that systems designed for high seismic risk are less vulnerable to gravity-induced progressive collapse and more robust than those designed for moderate seismic risk.

Motivated by a number of deficiencies in existing ductile fracture models for steel, a new micro-mechanical constitutive model is proposed. Damage mechanics principles are used and a scalar damage variable is introduced to represent micro-structural evolution related to micro-void nucleation, growth and coalescence during the ductile fracture process in steels. Numerical implementation and parametric studies are presented and discussed. Calibration and validation studies show that the proposed model can successfully represent ductile fracture of steels.

The models and simulation methodologies developed herein can be extended in future work to address the collapse resistance of three-dimensional models.