Failure Analysis of Fractured Gear Teeth on High Speed Shaft

The gear tooth of a high speed shaft cracked after servicing for two years. In this paper, the chemical composition, metallographic structure and fracture morphology of gear shaft materials were analyzed, and the fracture cause of gear teeth was determined. The result showed that there was large size nonmetallic inclusion in the gear teeth, which was result in the fatigue crack of the gear teeth.

Wind power growth gear box is one of the main components of wind turbine, which is arranged between wind turbine and generator.The wind turbine power is transferred to the generator to generate electricity, and the low speed of the wind turbine input is converted into the speed required by the generator.The service life of the gear is not only related to the material, thermal processing and heat treatment ‚ but also closely related to its operating environment.The gear is subjected to high torsional load, bending load and vibration load when rotating, so it has relatively high failure risk in the service.Common failure modes of gears include fatigue fracture, impact overload, surface contact fatigue and surface wear.Among them, the gear fatigue cracking is the main failure model of the gear shaft, and the fatigue broken teeth have a great impact on the normal operation of the gear box.At present, the commonly used gear shaft material is 18CrNiMo7-6, and the tooth surface is carburized.In this paper, The gear tooth of a high speed shaft cracked after servicing for two years.The gear shaft is made of 18CrNiMo7-6.This paper analyzes the failure cause of gear shaft tooth fracture base on the testing results.

Results of material chemical composition test
The chemical composition of the failed tooth was tested.The results are shown in Table 1.It can be seen that the chemical composition of the high-speed shaft material meets the requirements of the specification.

Analysis of macroscopic morphology
The macro morphology of the cracking tooth of the high-speed shaft is shown in Figure 1.There is a clear fatigue line on the fracture and the fracture has no obvious plastic deformation, which show that the fracture is fatigue cracking fracture.The crack initiation position is located at the waist of the tooth, with a vertical distance of 7.9mm from the top of the tooth and 2.5mm from the tooth surface, as shown in Fig. 1-(a) and Fig. 1-(b).After the fatigue crack propagated till the tooth flank fracture occurred under the action of alternating bending load.The final fracture area is very small, indicating that the fatigue load is not large and it is high cycle fatigue.

Analysis of micromorphology
The fracture sample was cut from the source area of the broken tooth, cleaned by ultrasonic wave, and observed under scanning electron microscope(SEM).The micro-morphology of the fracture is shown in Figure 2. The source area is slightly worn, and there are residues of non-metallic inclusions.Energy spectrum analysis shows that the non-metallic inclusions are mainly alumina inclusions.The non-metallic inclusions are distributed in a strip with a distribution length of about 2.8mm and the widest position of about 200μm.Clear fatigue striation can be seen in the steady propagation area.

Discussion
The above analysis results show that the chemical composition of the material used for the high-speed shaft meets the technical requirements, and no abnormality is found in the metallographic structure of the gear teeth.The grain size is 8.0.The carburizing layer on the tooth surface is evenly distributed, and the depth of carburizing layer is 1.7mm.No obvious manufacturing defects and external damage defects were observed on the gear shaft.There is abnormally large size (about 2.8mm long and 200μm wide) alumina inclusion in the fatigue initiation area.Generally, the influence of inclusion size on fatigue limit of steel is much higher than that of inclusion content.The fatigue limit of steel decreases gradually with the increase of inclusion size.The higher the strength value of steel, the more significant the change of fatigue strength caused by the change of inclusion size [1] .In this case, the total length of non-metallic inclusions near the surface of the gear teeth is more than 2.8mm, and the widest position is about 200μm, which seriously destroys the organizational connection of the material.Stress concentration effect accelerates the initiation of fatigue cracks.At the same time, due to the carburizing treatment of the teeth, the residual stress in the area near the carburizing layer is tensile stress, and the distribution of inclusions is located in the tension stress area.Finally, under the action of alternating bending stress, the tooth cracks and expands from the inclusion position, and finally cracks.
Studies [2] have shown that tooth flank fracture is related to tooth stress state, tooth flank geometry, local material strength and residual stress.Since the fracture point of tooth surface is usually below the hardened layer, the influences caused by tooth surface friction, thermal stress and tooth surface roughness can be ignored.Based on a large number of theoretical and experimental studies, FZG has developed a calculation method, which connects the shear stress in the depth direction under the Hertzian load of the tooth surface with the local material strength (directly derived from the local hardness curve), so that the local material response of any volume unit under the tooth surface can be calculated [3] .The method proposed by ISO is based on the research results of FZG.The local exposure value(AFF) is calculated by considering various stresses, including residual stress, to evaluate the fracture risk of the tooth surface of the gear teeth [2,4,5] .In the formula (1), the hardness distribution and residual stress distribution should be tested, and the local equivalent stress τeff and local material shear strength τper were calculated by combined with applied load and tooth surface geometry.It can be seen that the peak value of AFF is located at the position of 2bH, and bH is the Hertz contact half-width, indicating that the greatest risk point is located inside the tooth surface, showed in the Fig. 4. Generally, the fatigue fracture risk of tooth surface is high when the AFF value exceeds 0.8.In this case, there are large-size inclusions at the rack fatigue source, and the local stress concentration caused by them cannot be ignored.Murakami et al. [6] put forward the inclusion equivalent projected area model, using the inclusion parameter (S is the projected area of the inclusion perpendicular to the stress axis plane) to explain the fatigue behavior of non-metallic inclusions in high-strength steel.The critical diameter of the non-metallic inclusions inside the material decreases with the increase of core hardness, which mean that the higher the hardness of the material, the greater the risk of fatigue cracking caused by inclusions.Appropriate reduction of core hardness can reduce the risk of fatigue cracking caused by inclusion, and similar conclusions have also been discussed [7] .However, reducing the strength of the core will reduce the ability of the tooth to resist the bending fatigue cracking at the root position.Therefore, it is necessary to combine the design load, the depth of the hardened layer and the size control of inclusion in the material to evaluate comprehensively.At present, there are limited reports on in situ characterization of inclusions using micro confocal Raman spectroscopy [8,9] .In situ analysis of these inclusions was carried out in this paper.The experiment used micro confocal Raman spectrometer (NCS Testing Technology CO., LTD), selected 532nm laser wavelength, scanning range of 50cm -1 ~8000cm -1 .In Figure 5, Raman displacement of 4358cm -1 is the characteristic peak of alumina inclusion, which further verifies the conclusion of SEM and metallographic analysis.(2) Tooth surface cracking prevention needs to consider the depth of carburizing layer, matrix hardness, tooth load and inclusion size.

Figure 1 .
Figure 1.Morphology of the gear shaft (a) Morphology of the crack shaft; (b) The location of the cracking initiation

Figure 2 .Figure 3 .
Figure 2. Micro-morphology of the fractures (a), (b)Cracking initiation of fracture; (c) Content distribution of Al in the EDS testing area; (d) Content distribution of O in the EDS testing area; (e) Fatigue striation in fracture

Figure 4 .
Figure 4. Relation between local equivalent stress τeff, local material strength τper, local material exposure value AFF and depth

Table 1 .
The