Condition monitoring of railway axles based on low frequency vibrations

Railway axles are safety-critical components and their structural integrity needs to be properly monitored against potentially catastrophic failures. This paper proposes a method for the continuous condition monitoring of railway axles by measuring the axle in-service bending vibration and diagnosing the presence of a fatigue crack based on the examination of harmonic components with periodicity corresponding to an integer sub-multiple of the axle revolution period (nxRev). To analyse the feasibility of the proposed method, full-scale measurements were performed on cracked and non-cracked railway axles undergoing fatigue tests and a mathematical Finite Element (FE) model of a cracked railway axle was defined and used to reproduce the laboratory experiments and also to analyse the case of a railway wheelset with cracked axle rolling on a railway track. The results of both experimental and numerical investigations show that the 1xRev, 2xRev and 3xRev components of axle bending vibration can be used to detect the presence of cracks in axles, provided the crack area is in the range of 16% of axle cross section or larger. Hence, the proposed method cannot replace periodical non-destructive inspections, but may serve as an additional safety measure detecting cracked axles in an advanced stage of the damage process but still on time to avoid in-service failures. (C) 2015 Elsevier Ltd. All rights reserved.