Fretting fatigue in steel bolted connections

The steelwork in a potash mineshaft is subjected to repeated lateral loads due to the lateral motion of the cages and skips that transport personnel, ore, and equipment up and down the mineshaft. As a result, the steel assemblies, including their bolted connections, must be designed to prevent failure due to fatigue. However, due to uncertainty regarding the fatigue behaviour of the connections, designers must take a very conservative approach, which could result in an uneconomical design. The main objective of this research was to determine the fatigue behaviour of slipcritical bolted connections when different bolt types and surface finishes are used. A325 high strength bolts and C50LR Huck tension control bolts were used as the different bolt types. As-received mill scale steel plates (Class A surface finish) and blast-cleaned surfaces with a Cathacoat 302HB coating (Class B surface finish) were used as the different finishes. A digital image correlation system, as well as optical and scanning electron microscopic examination were used to characterize the modes of failure of the specimens. Bolted connections assembled with the Class A surface finish failed due to fretting fatigue damage and crack initiation took place some distance away from the hole in a partial slip region between a stick region adjacent to the bolt hole and a gross slip region further from the hole where the relative motion between the plates was highest. On the other hand, specimens with the Class B surface finish failed due to bending fatigue caused by the eccentricity between the tension forces in the plates in the single-lap bolted joints, and crack initiation took place at the hole edge where the stress concentration was higher. The bolt type did not have a large effect on the fatigue behaviour, except that the tension in the tension control bolts may have been slightly higher, resulting in a slight improvement in the fatigue life. In general, the fatigue life results were lower than those in the S-N curve given in CSA S16-14, differing from the standard curve by an increasing margin as the stress range increased due to the effects of bending within the specimens