New insight into the fatigue-atmospheric corrosion process of A1N steel for railway axles

In this paper a detailed investigation about the early phases of the corrosion fatigue process of this material is presented. The corrosion pits appears in the early stage of the corrosion fatigue life and a preferential site of pit nucleation has been observed at the ferrite-ferrite grain boundary. At the bottom of these primary pits has been observed the formation, due to electrochemical action of a secondary pit that works as a trigger for the pit-to-crack transition. The growth of micro-cracks has, also, been analyzed and results dependent on the stress level and strongly affect by the coalescence phenomena with other growing cracks. Key points for the development of a new model for the corrosion-fatigue life prediction of railway axles are defined. Introduction The design of railway axles is performed for infinite fatigue life related to loads level depending on the railway vehicle type, and did not usually consider different in-service degradation and damage mechanisms such as environmental corrosion and corrosion fatigue [1, 2]. In recent years, the degradation due to corrosion shown by railway axles has increasingly become an area of concern. Some references report cases of axle failure due to crack propagation from corrosion pits. Hoddinott [3] reports that about five mid-span failures of in-service axles occurred in the UK from 1996 to 2003, with four being connected to the presence of diffused axles surface corrosion and corrosion pits. On the other side of Atlantic, the Transportation Safety Board of Canada [4] reported that, in one case, axle failure was caused by corrosion pits under the journal bearing. It also mentioned that another seven similar failures occurred between 1998 and 2000. Among the different types of corrosion damage, pitting is one of mechanism in triggering widespread fatigue crack initiation and reducing fatigue life of the material. Hidden corrosion pits are also difficult to detect nondestructively. Previous research [5, 6] has provided some features about the effects of atmospheric corrosion on fatigue properties of A1N steel. Under the interaction of cyclic load and corrosive environment cyclic load facilities the pitting, and corrosion pits, acting as geometrical discontinuities, lead to nucleation of a consistent number of small cracks. These small cracks are then able to cross the microstructural barriers with ease and at a much faster growth rate than in air. These effects result in a considerable decrease in the fatigue life of the material and in the disappearance of the ‘‘knee” of the S–N diagram. In this paper a detailed investigation about the early phases of the corrosion fatigue process of this material is presented. The corrosion pits appears in the early stage of the corrosion fatigue life and a preferential site of pit nucleation has been observed at the ferrite-ferrite grain boundary. At the bottom of these primary pits has been observed the formation, due to electrochemical action of a secondary pit that works as a trigger for the pit-to-crack transition. The growth of micro-cracks has, also, been analyzed and results dependent on the stress level and strongly affect by the coalescence phenomena with other growing cracks. When the crack length became greater than 1 mm the crack growth rate tends to the crack growth rate in air. On the basis of the detailed analysis of the different stage of the corrosion fatigue process, a discussion about application of the corrosion fatigue models present in literature to the life estimation of railway axles is presented and key points for the development of a new predictive model are defined. Material and experimental procedures Material. A1N is a normalized 0.35% carbon steel, widely used in the manufacture of railway axles: the matrix consists of a ferritic-pearlitic microstructure with a 20–40 m ferrite grain size. Its basic mechanical properties are: ultimate tensile strength UTS = 597 MPa and monotonic yield strength y,monotonic = 395 MPa. Cyclic properties are as follows: 0.2% cyclic proof stress y,cyc0.2 = 357 MPa, 0.05% cyclic proof stress y,cyc0.05 = 289 MPa. The parameters of the cyclic Ramberg– Osgood relationship are equal to Ecyc = 209,303 MPa, n = 0.150395 and H = 907.34 MPa. Specimens Specimens are hourglass shaped with a minimum diameter of 10 mm. Following machining, the specimens were polished up to #1000 grit emery paper and then mirror polished with diamond paste. Corrosion-fatigue tests. In order to investigate the evolution of the corrosion fatigue damage of A1N steel a series of corrosion–fatigue tests were run at R = -1, using a four point rotating bending machine (capacity of 35 Nm) working at a frequency of 10 Hz. This frequency value is representative of axles during a service speed equal to 100 km/h. Corrosion was continuously applied to specimens by means of a dedicated dropping system with a dripping flow rate of 40 cc/min (Fig. 1a). The artificial rainwater solution [7], characterized by pH = 6, was formulated as follows: ammonium sulfate 46.2 mg/dm 3 , sodium sulfate 31.95 mg/dm 3 , sodium nitrate 21.25 mg/dm 3 and sodium chloride 84.85 mg/dm 3 . The corrosion–fatigue behavior of the A1N steel was also characterized by freely corroding condition. Corrosion potential is monitored according to the diagram shown in Fig. 1b [8]. Tests were carried out at stress levels S in the range of 180–400 MPa that include both the EN13103/4 design limit (design = 332 MPa), and the BASS design limit (design = 220 MPa), which is valid for steels with UTS of 550–650 MPa, and corresponds to a median corrosion–fatigue life of 4 10 7 cycles. This range, below the fatigue limit in air of the material (S = 510 MPa) shows a finite life in presence of atmospheric corrosive environment which eliminates the fatigue limit of the material, as can be observed in Fig. 2. Tests on smooth specimens has been interrupted, and not restarted, at different stages (5÷50%) of the corrosion fatigue life, as estimated by the SN diagram of Fig 2a.