| Résumé | This article describes the failure analyses of a 30”-diameter cast iron main that failed suddenly at the bell–spigot joint in Cleveland, Ohio in March 2008. The main had been operating largely without problems since its installation in 1880. The failure was analyzed using three models developed previously, each of which considered a specific failure mechanism candidate, namely (a) operational load or static analysis, (b) joint failure due to ground movement (settlement), and (c) fatigue failure. Operational load analysis clearly showed that while the factor of safety was significantly reduced in the presence of a concrete vault installed in 2002 just above the pipe, the margin of safety was still sufficient for the main to have performed safely. Ground movement analysis of the jointed pipe showed that the construction of the concrete vault above the joint of the 30”-cast iron main subjected the joint to rotation, which might have ultimately cracked the bell leading to failure. Under this circumstance, two failure scenarios are possible: namely, (a) additional rotation was large enough to cause the bell to split, or (b) additional rotation was sufficient to induce a minor crack in the bell but fatigue (repeated) loading caused the crack to grow over time until the combination of crack length and loading was sufficient to cause the bell to eventually split. The color differences along the fracture surfaces indicated that the fracture occurred in two distinct stages. This observation suggests that the second scenario is more plausible than the first. |
|---|
