Surface fatigue initiated transverse defects and broken rails -- an International Review

Report, 2014

The current report briefly compares some operational experience of cracked and broken rails from China, Russia, South Africa, Sweden, UK and USA. Four key questions are addressed: 1. Is the critical crack length, i.e. the length of a surface initiated crack that causes a rail break (reasonably) constant in an international perspective? 2. Is it (reasonably) constant over a line? 3. Can the depth when a rolling contact fatigue crack deviates to a transverse propagation be estimated? 4. Is it (reasonably) constant in an international perspective? The answers can briefly be summarized as 1. No. Deviations in crack sizes from roughly 10% up to roughly 80% of the railhead area at fracture have been found. 2. Not generally, but for some lines this seems to be the case if fractures at the same season are considered (i.e. climate effects are excluded). 3. There are indications that this depth is in the order of 5 mm with a fair amount of scatter. However it is very difficult to identify from a photo whether an area of the fracture surface actually corresponds to inclined fatigue crack propagation. 4. With a reservation in the considerable scatter, there seems to be some consistency also in an international perspective. Details on how these conclusions were reached are given in the report.