Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Civil Engineering

Committee Chair

Robert J. Connor

Committee Member 1

Mark D. Bowman

Committee Member 2

Amit H. Varma

Committee Member 3

Matthew H Hebdon


The objectives of this research were to determine if mechanically-fastened steel built-up axially-loaded tension members are resistant to complete cross-sectional fracture when a single component suddenly fractures and if so, to develop simplified methods of analysis for post-fracture load redistribution behavior for calculation of remaining fatigue life. The ability to resist complete member failure and perform at a predetermined level of reliability with a failed component is referred to as internal member redundancy (IMR). The experimental program included seven full-scale tests; five were extremely demanding fracture tests, including one specimen with a tack weld placed in the path of a running fracture, and the other two were full-scale specimens removed from a 1940’s era built-up riveted deck truss, which were partially failed and tested to calibrate finite element models. Comprehensive finite element model-based parametric studies were carried out investigating the post-fracture load redistribution behavior and resulting stress amplification for multi-component, angle-only, and two-channel type axial members. Simplified closed-form solutions were developed for members of any geometry that fall within these broad categories of built-up axially-loaded tension members.

The IRM evaluation process is intended to combine the concepts related to probability of detection (POD) of fatigue cracks on steel bridges with the fatigue damage tolerance of mechanically-fastened built-up members. Preliminary results from ongoing POD research at Purdue University suggests that the steel bridge industry has unknowingly been relying on internal redundancy of mechanically-fastened built-up members. In other words, due to the difficulty of finding small cracks in components of built-up members, it is more likely that traditional arms-length inspections of fracture-critical members will find broken components instead. More candidly stated, this research provides the industry with quantitative analysis for the purpose of establishing rational inspection intervals for built-up axial members that are realistic about what can be reliably found during inspections and for what duration undiscovered damage can be safely tolerated due to internal member redundancy of the mechanically-fastened built-up member.