Date of Award

Fall 2011

Document Type


Degree Name

Doctor of Philosophy (PhD)


Civil Engineering

First Advisor

Christopher M. Foley

Second Advisor

Stephen M. Heinrich

Third Advisor

Philip Voglewede, Baolin Wan


Robustness is a desirable property in any structural design. Robustness may be thought of as the building's inherent structural ability to resist loads other than those anticipated during design. Examples of unanticipated loads are damage to a column or beam due to impact; damage due to the concussion of a blast; strength reduction due to extreme heat during a fire; and localized accidental overload of a beam or column. Each of these events, although very different in their nature, has one major point in common; in each case the path of load resistance changes in a way that was not and perhaps could not be anticipated during the original design.

A three-story, pre-Northridge Boston building was chosen as the basis for this study. Although the building is generic, it is representative of typical office buildings constructed in non-seismic areas of the Central and Eastern United States.

The building was modeled using structural engineering software capable of performing a second-order, inelastic analysis with user defined connection characteristics. Ten connection models with varying parameters were considered for evaluation and comparison of their capabilities to withstand loading after a column has been rendered ineffective. The varied parameters include connection placement; bolt quantity; inclusion or exclusion of a seat angle; inclusion or exclusion of the tensile capacity of a reinforced composite floor slab; and depth of concrete in the composite floor slab. Both two-dimensional and three-dimensional frameworks are evaluated.

A robustness measure was defined as the relative change in applied load ratio compared to the applied load ratio of a base system. The base system was identified as one that would be the result of a design considering efficiency in strength and serviceability to withstand code required load combinations.

Models that include both a seat angle and tensile capacity of slab reinforcement produced the most robust system. Adding one additional bolt produced only modest improvement in robustness. Adding depth of concrete was counterproductive due to increased weight. A three-dimensional analysis is very important to adequately quantify robustness in a particular framework. The opposing girders and beams along the gridlines, along with the infill framing, work together to resist unanticipated loading and their total contributions are significant.