Since the railway rail was subjected to cyclic loading and served about 6 years, it is rational to consider that this railway rail might be failed due to fatigue, even in giga cycle fatigue regime [5–7]. The arc boundary of fan-shaped area looks like a beach mark when observed macroscopically, as seen in Fig. 2(c). Then it can be deduced that the crack origin might be at the corner of darkly fan-shaped area (viz., the small bright spot as shown in Fig. 2(c)). However, taking into account that the small bright spot was next to darkly fan-shaped area, it is obviously to deduce that the small bright spot would not be the crack origin. The beach marks which were the classical features of metal fatigue were not observed from the macroscopic observations (the arc boundary of fan-shaped area is actually not a beach mark, we will discuss that hereinafter). However, the chevron patterns can be clearly observed at the fracture surface of rail bottom (Fig. 2(d)). Therefore, we can deduce that the crack origin should be at the tip of chevron patterns (area 1), and the crack growth direction is along the diverging direction of river patterns, as shown in Fig. 2(d). It can be deduced from the flat fractography and chevron patterns that the macro fracture feature is brittle fracture.
3.2. Chemical analysis
The sample used for chemical analysis which was sampled from railhead was analyzed by ZSX Primus II X-ray fluorescence spectrometer. The results were shown in Table 1. It was demonstrated that the chemical compositions of the rail steel were in accordance with the standard of P60U75V [8]. Therefore, the compositions of the rail steel were normal.
3.3. SEM observations
The fracture surface at the rail bottom, viz. surface S1 (as shown in Fig. 2(c)), was cut from the failed railway rail, and then cleaned by alcohol and dichloroethane. After that, surface S1 was observed by ZEISS-SUPRA 55 field emission scanning electron microscope (FESEM) in detail. The morphology inside the darkly fan-shaped area is shown in Fig. 3(a) which shows a relatively flat surface. The qualitative chemical compositions of this area are analyzed by EDX, also shown in Fig. 3(a). Higher oxygen contents were detected, which demonstrates that this area was oxidized heavily. The typical morphology in area 1 is shown in Fig. 3(b) which shows the typical feature of cleavage fracture. The fan-shaped patterns, cleavage step and river patterns which are the typical feature of cleavage fracture are observed in this figure. Fatigue striations which were the typical microscopic features of metal fatigue were not observed in the fracture surface. The micro fractography of the chevron patterns area is shown in Fig. 3(c) which is similar to Fig. 3(b). The fracture surface outside the darkly fan-shaped area is clean and fresh, almost no oxygen is detected. Combined with the experimental results outside and inside the darkly fan-shaped area, it can be deduced that the darkly fan-shaped area might be an incomplete fusion area during welding.
3.4. Metallurgical observations
Firstly, surface S2 (as shown in Fig. 2(c)) was polished to observe the distribution of inclusions. It was shown in Fig. 4 that some bigger slag inclusions with sharp angular shape were observed at surface S2 close to the fracture surface at the rail bottom. The size of these slag inclusions was about at least 126 mm defined by Murakami’s effective projective area model [9]. EDX demonstrated that the composition of these slag inclusions was alumina. After etched by 3% nital the metallurgical
structure of surface S2 close to the fracture surface was observed by optically metallurgical microscope (OMM). Continuous ferrite networks and pearlite colonies were observed, as shown in Fig. 5(a). It was also demonstrated from Fig. 5(a) that the size of pearlite colony, viz. the area surrounded by continuous ferrite networks, was rather heterogeneous. The biggest size of the pearlite colony was about 726 mm in diameter, the smallest size 68 mm. After making an obvious mark at the crack origin site (area 1) on surface S1, surface S1 was polished and etched by 3% nital in order to observe the metallurgical structures. The metallurgical structures in area 1 were pearlite, continuous ferrite networks and a mass of ferrite fragments distributed
inside the pearlite colonies, as shown in Fig. 5(b). Considering the weaker strength of ferrite distributed like nets compared with pearlite; it can be deduced that the crack might be initiated from the ferrite networks.
4. Discussion and analysis
As introduced in Section 1, the railway rail was mainly subjected to cyclic loading. In this case study, the macroscopic beach marks and microscopic fatigue striations were not observed at the fracture surface. In addition, the typical chevron patterns were observed. And the feature of cleavage fracture was observed at the tip of chevron patterns. Given all of that, we can draw a conclusion that the railway rail is mainly caused by overload even though it is subjected to cyclic loading. Considering the abnormal metallurgical structures (ferrite networks distributed along the grain boundaries) at the crack origin, the crack is supposed to be initiated from the weaker ferrite networks which are caused by welding. Therefore, it is much needed to eliminate the ferrite networks by improving the welding technology.
4.1. Stress analysis
In this case study, this railway rail was mainly subject to alternate bending stress due to vehicle weight, as shown in Fig. 1. It is shown from Fig. 1 that the rail head was subject to compressive stress and the rail bottom was subject to tensile stress.
Therefore, the rail bottom was subtle to fail from the point of view of mechanics. The rail bottom can be approximately considered to subject fatigue load in the longitudinal direction of railway with R = 0 (R was stress ratio), and the applied stress at the lower bottom of rail was approximately the maximum. In view of the heavily incomplete fusion area (darkly fan- shaped area in Fig. 2(a)) at the bottom corner of rail bottom, the crack was supposed to be initiated from this incomplete fusion area. Nevertheless, in fact it is not the case. The residual stress must be considered. In addition to the residual tensile thermal stress due to track installation and temperature (as shown in Fig. 1), the welding residual stress was also of great importance. The welding residual stress was usually detrimental; therefore, many rail failures were initiated from the weld [4]. The residual stress and applied stress due to the train’s gravity can be superimposed together. The superimposed stress will induce the failure of the rail under certain conditions.
4.2. The role of fatigue
The railway rail is failed by overload; however, we cannot completely deny the role of fatigue. The damage trace generated by cyclic loading was not observed at the crack growth area; however, the fatigue damage was almost inevitable by taking the longer service life into account (about 6 years). On the other hand, usually the fatigue damage cannot be easily detected due to the complicated working situations of a component.
4.3. Suggestions
The failed railway rail was caused by overload. The crack origin was the ferrite net induced by inadequate welding technology. Therefore, in order to prevent similar failures in future, the welding process must be improved. For example, pro and postweld heat treatments should be conducted to eliminate the bigger slag inclusions and ferrite networks along the grain boundaries. On the other hand, it is of great importance to control the load of train.
5. Conclusions
This failed railway rail is caused by overload. The crack is initiated from the weaker ferrite networks which are induced by inadequate welding technology. It is of great importance to improve the welding technology, and control the load of train in future.
Acknowledgments
This work was financially supported by National Natural Science Foundation of China (Grant no. 51101094) and Fund project of Technology Development of Shandong Academy of Sciences (Grant no. KJHZ 2011-04).
Sedang diterjemahkan, harap tunggu..