Abstract
As unpredictable fracture of the front axle will lead to serious safety accidents, its failure behavior study is essential. This paper analyzes typical cracks on I-shaped and leaf springs of the front axle made by Steel A and Steel B. Steel B has a lower content of S, N, Cu, and Sn than Steel A. In order to identify the failure mechanism difference, the low magnification etch test, and metallographic and scanning electron microscope analyses were conducted. The results show that the microstructures are tempered sorbite, and the cracks are with intergranular fracture characteristics. The aggregated distribution of sulfide inclusions in the matrix of Steel A is higher than that of Steel B. Copper-rich particles are detected in the cracks of Steel A. The result indicates that the purity distinction of raw steel can have a big difference on the cracking rate of the front axle even under the same manufacturing process. Relevant suggestions to reduce the cracking are put forward from the aspects of raw materials and process conditions.
1. Introduction
Front axle is one of the vitally important parts of an automobile. It needs high strength, stiffness, and good fatigue resistance, and is mainly subjected to a combination of shear stress and bending stress [1]. Most of the failures are due to crack growth generation caused by fatigue load, which then propagates and leads to failure. Because of the unpredictability, fracture failure of the front axle will lead to serious safety accidents [2, 3]. Therefore, it is significant to study the failure mechanism of automobiles’ front axle to reduce the occurrence of accidents.
Different materials have different strength, hardenability, impact toughness, and fatigue performance, which significantly influence the failure behavior. The front axle is commonly manufactured using materials 42CrMo, 40Cr, and other micro-alloyed steels. Several studies have been conducted with regard to the effects of the front axle material. Wu et al. put forward the composition design method and empirical formula for the development of air-cooled bainite micro-alloyed steels for front axle beams of heavy trucks [4]. The results showed that austenitic transformation can be controlled by selecting the chemical composition of the front axle to obtain the mechanical properties for specific applications. Dhanasekaran et al. indicated that the feasibility of vanadium-alloyed steels in automobile front axle applications [5]. The results showed that vanadium-alloyed steels with normal cooling conditions meet the critical requirements of heavy vehicle front axle application. Mao et al. investigated excellent characteristics of the nitriding layer of 42CrMo steel, which can effectively improve the wear resistance under heavy loads [6]. A constitutive model is proposed by Zhu et al. to describe the dynamic mechanical behavior of 4z2CrMo steel [7]. Hot cracking at circular welding joints of quenched and tempered 42CrMo steel was studied by Zhang et al. [8].
As well known, the microstructure of the material directly affects its mechanical properties. Therefore, it is of great significance to improve the microstructure of the material through reasonable processing and heat treatment technology for enhancing the quality of the front axle [9]. The processing technique of automobile front axle has been widely studied. A 3D thermo-mechanical coupling finite element model was established by Wang and Hua to predict the microstructure evolution in the automotive front axle beam during the hot forging processing [10]. Tawancy and Al-Hadhrami analyzed a section of the fractured rear axle shaft made of induction-hardened steel, which was removed from the overturned automobile, to determine the most probable reason of failure [11]. According to microstructural characterization and microhardness measurements, the results showed that the shaft was improperly heat treated resulting in a brittle case. Cai analyzed the kinematic relation between the surface of the rolling die and the forging piece during the roll-forging process, and then described the roll-forging manufacturing technique for automobile front axles [12]. An accurate 3D FE model of roll-forging of automotive front axle beam is developed by Zhuang et al. under the Deform-3D platform [13]. Kannan and Srinivasan found that the final turning process step is detrimental to fatigue life, whereas the sequence with induction hardening and induction tempering as the final process step produces least tensile surface residual stress, and hence improves the fatigue life [14]. According to the characteristics, technological conditions, and requirements of automobile front axle forgings, Li et al. analyzed the processing difficulties of automobile front axle [15]. Hawryluk and Jakubik used a variety of tools and special control systems to design and optimize the entire forging process [16].
Corrosion, wear, and fatigue are the main causes of failure of mechanical parts. The main failure form of the front axle is fatigue damage. In addition to chemical composition and processing technique researches, many scholars also focus on the failure mechanism itself. Guo et al. analyzed the possible reasons for the fracture of a welded front axle tube structure from a mini truck [17]. Ognjanovic et al. made an empirical analysis on fractures and fracture processes of the traction shafts and axles [18]. Clarke and Halimunanda provided a fracture mechanic approach to determine when the axle failed and then analyzed the rollover accident [19]. Hu et al. attempted to evaluate the remaining life and the critical non-propagating crack depth of induction-hardened S38 C railway axles with an assumed surface crack [20]. FEA was used by Zheng et al. to analyze the cause of the cracking on the frame of the wide-body mining dump truck, and the design of the frame was also improved by them [21]. Topac et al. determined fatigue crack initiation locations and minimum number of load cycles before failure initiation, and then proposed an improved design scheme to increase the fatigue life of the housing [22]. Besides that, Topaç et al. also analyzed the failure modes and design improvement of an eccentrically loaded connecting rod for a double front axle steering linkage prototype [23]. Zhang et al. analyzed the fracture response of a pipeline girth weld with surface or embedded elliptical cracks subjected to large plastic bending using finite element simulations and established two failure assessment diagrams to perform fracture assessment on the specific pipeline [24]. Hoh et al. investigated the Dugdale crack behavior near a circular inclusion under uniaxial tension and proposed the influence of the crack–inclusion distance, materials’ constants, and load-to-yield stress ratio on the PZS and CTOD [25]. Besides the studies discussed above, some other relevant examples of studies conducted on this topic can be found by Bayrakceken [26], Shao et al. [27], and Xu et al. [28].
Front axle failure can be caused by a number of factors. It may be the defects of raw materials or any influences of the processing technology. It is of great significance to find out the cause of fracture failure and prevent similar occurrences. However, few researches have focused on the purity effect of high-strength steel on the failure behavior of front axles. This paper analyzes typical cracks on I-shaped and leaf springs of front axles made by Steel A and Steel B, respectively. Chemical composition comparison, low magnification etch test, metallographic, and scanning electron microscope analyses were conducted. The crack characteristics, comparison of the two kinds of steel, and intergranular cracking mechanism are emphatically discussed, and relevant suggestions to reduce the cracking are put forward from the aspects of raw materials and process conditions.
2. Materials and Samples
The blank materials of automobile front axles are Φ140 mm round bar of high-strength alloyed structural Steel A and Steel B. Principal chemical element analysis of Steel A and Steel B was performed using the ULTIMA 2C ICP direct reading spectrum of Company JY in France. C and S were determined by the EMIA-820V infrared carbon–sulfur analyzer of Company HOEBA in Japan. O, N, and H were investigated by the EMGA-600WSP combined equipment. Trace analysis of the five harmful elements was carried out by XseriesII ICP-MS (Inductively Coupled Plasma Mass Spectrometry) of American thermoelectric company. The results are shown in Table 1.
By comparing and analyzing the measured results in Table 1, it can be concluded both Steel A and Steel B meet the requirements of alloyed structural steel in GB/T 3077-2015 [29] for the content of C, Si, Mn, Mo, and Cr. However, Steel B has lower contents of S, Cu, N, and Sn, which indicates that Steel B has a higher purity than Steel A.
The production process of the front axle is as follows: Φ140 mm Steel A or B round bar ⟶ blanking ⟶ multiple roll-forging molding at 1200°C temperature ⟶ scrap edge ⟶ hot straightening ⟶ air cooling ⟶ thermal refining (water quenching after heating 850°C for 110 min ⟶ tempering 590°C for 150 min) ⟶ shot peening ⟶ crack detection.
The shape of the front axle is complex and its section size varies greatly. The maximum section thickness is a square of about 70 mm high and the minimum section thickness is only 14 mm. In order to ensure the mechanical properties of the front axle requirements, its surface hardness should be range from 276 to 326HV, the core hardness should be greater than 276HV, and the tensile strength should range from 880 to 1030 MPa.
Cracking phenomena were both found in the front axles after quenching and tempering heat treatments. The cracking rate of Steel A is obviously higher than that of Steel B. According to data statistics, the cracking rate of Steel A is about 15%, and that of Steel B is about 2%.
Cracked components were inspected on-site and the cracked parts of the axles were randomly investigated. As Figures 1–3 show, 7 samples were taken from Steel B and 27 samples were taken from Steel A. Cracks were found on the upper, lower, and sides of the front axle, and cracks were distributed irregularly. Typical crack samples at the I-shaped position and leaf spring position were selected from the above samples of Steel A and Steel B. In order to identify the failure mechanism difference, chemical composition comparison, low magnification etch test, metallographic, and scanning electron microscope analyses were conducted. Sample numbers are shown in Table 2.
3. Experiment Analysis
3.1. Low Magnification Etch Analysis
Nine samples in Table 1 were used in the low magnification etch experiment. The samples were etched and polished with 10% hydrochloric acid; the low magnification structure observation results are shown in Table 3.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
The cross-section samples of low magnification etch analysis are shown in Figure 4. Microstructural defects were not observed on the low power-etched surface of these nine samples, such as white flake, skull patch, air bubble, shrinkage cavity, etc. Five cracks A, B, C, D, and E can be seen on the low magnification pickling surface of sample X-3 and Y-3-1. Except crack A at the bottom of the leaf spring, the other four cracks all exist in the R-angle position of the front axle. All of the cracks are different in depth. In sample X-3, the crack depth at A is about 18 mm, the crack depth at B is about 4 mm, and the crack depth at C is about 8 mm. In sample Y-3-1, the crack depth at D is about 10 mm and the crack depth at E is about 1 mm. The sample has been deformed by roll-forging for many times. The grade in Table 3 is evaluated by GB/T1979-2001 [30].
3.2. Metallographic Analysis
3.2.1. Crack Observation
Crack samples were taken from the cracking parts of X-1, X-3, Y-1, and Y-3 as shown in Figures 2 and 3 for metallographic observation. The macroscopic morphology of the crack surface is shown in Figure 5. After inlaying, grinding, and polishing, the crack characteristics of the samples are shown in Figure 6. The cracks are different in depth as shown in Table 4.
After the samples were etched in 3% nitric acid alcohol solution, the microstructure characteristics were presented in Figure 7. The crack observation results show the crack defects are irregularly distributed on the upper, lower, and sides of the front axle. However, all cracks propagated longitudinally along the front axle. The macroscopic characteristics of them presented thin and straight single strips. No transverse crack, network crack, or dendritic crack distribution was found. The aggregated inclusions were not observed near the crack. The crack is thick in the middle and thin at both ends. Some small intergranular cracks were observed on both sides of the main cracks. The end of the cracks is tortuous, tapering, and has intergranular propagation characteristics. Especially, the near surface of the cracks showed a “beak-shaped” characteristic with an angle of 45° to the surface and the typical example is shown in Figure 6(c). The crack propagated from the middle to both ends. It is obvious that there was iron oxide coating in the crack, especially on both sides of the crack tail. There is no abnormity in the microstructure after etching, all of them are tempered sorbite, and there is no oxidation and decarbonization on both sides of the crack.
3.2.2. Inclusion Test
Longitudinal samples were taken from the uncracked parts (X-2, X-4, Y-2, and Y-4) of Steel A and Steel B, and inclusions were observed on the grinding metallographic polishing surface. The results are shown in Table 5 and Figure 8. The inclusion type and grade in Table 5 are evaluated by GB/T10561-2005 [31].
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
(k)
(l)
It can be seen from Table 5 that the sulfide inclusions in the round bars supplied by Steel A have an aggregated distribution with the highest grade of 2.5 (Figure 8(c)), while the sulfide inclusions of Steel B have a dispersed distribution with the highest grade of 1.5 (Figure 8(k)).
3.2.3. Microstructure Observation
After etching with 3% nitric acid alcohol solution, the metallographic microstructure of the above four samples was all tempered sorbite and had a normal quenched and tempered state structure. The microstructure of samples X-4 and Y-2 had obvious banded characteristics as shown in Figure 9.
3.3. Scanning Electron Microscope Analysis
The cracked parts of X-1, X-3, Y-1, and Y-3 as shown in Figure 5 were cut into small pieces. These small pieces were brittle failure at liquid nitrogen temperature to obtain cross-section samples. Then, they were cleaned with an ultrasonic cleaning machine and analyzed by Quanta 400 scanning electron microscope and energy dispersive spectrometer. The results are as follows:
The microscopic morphology of the Y-1 crack surface is shown in Figure 10. The crack surface is seriously oxidized, and the energy spectrum analysis is shown in Figure 10(c). After the oxidation layer was removed by a rust remover, the microscopic morphology of the crack surface was shown in Figures 10(d) and 10(e), and the intergranular characteristics are more obvious. Particularly, it can be seen from Figure 10(d) that there are plastic deformations and dimples in the proximal surface of the cracked surface. As described in Figure 10(f), fresh fracture after brittle fracture is characterized by cleavage, and no obvious inclusions were found at the cleavage section. The crack part was cut out and the section was ground and polished. Then the local morphology of the crack was observed under an electron microscope as Figure 10(g) shows. Iron oxide is embedded in the crack, and the energy spectrum analysis is shown in Figure 10(h). As shown in Figure 11, the observation results of the Y-3 crack surface are similar to those of Y-1.
(a)
(b)
(a)
(b)
The macroscopic morphology of the X-1 crack after cutting is shown in Figure 12(a). The microstructure of the crack surface is shown in Figure 12(b), and the surface is also seriously oxidized. The energy spectrum analysis is shown in Figure 12(c). After the oxidation layer was removed by the reagent, the microscopic morphology of the crack surface was shown in Figure 12(d) and Figure 12(e). The intergranular characteristics are more obvious, and it can be seen from Figure 12(d) that there are plastic deformations and dimple characteristic in the near-surface layer of the cracks. It is shown in Figure 12(f) and Figure 12(g) that fresh fracture after brittle fracture has a cleavage characteristic, and many strips of MnS inclusions are found on the cross section. Energy spectrum analysis is shown in Figure 12(h). The crack part was cut out and the section was ground and polished. Then the local morphology of the crack was observed under an electron microscope as shown in Figure 12(i). As described in Figure 12(j), iron oxide is embedded in the crack, and small white copper-rich particles can be seen locally in the oxide layer. Energy spectrum analysis of iron oxide is shown in Figure 12(k). Energy spectrum analysis of copper-rich particles is shown in Figure 12(l) and Figure 12(m). As shown in Figure 13, the observation results of the X-3 crack surface are similar to those of X-1.
(a)
(b)
(c)
(a)
(b)
(c)
In summary, the main sections of the cracks are all intergranular fracture, and the intergranular section is seriously oxidized. No more valuable morphology and composition information has been detected after rust removal. Plastic deformation and dimple characteristic are in the near-surface layer of the cracks, indicating that it was finally torn. This is consistent with the observation that the cracks are thin at both ends and thick in the middle on the metallographic grinding surface and the near-surface layer is “beak shaped.” Both of them indicated that the intergranular crack propagates from the middle to both ends. There are more strips of MnS inclusions on the low temperature brittle section of the sample of Steel A, which is also consistent with the observation results of metallographic grinding surface inclusions. The low temperature brittle section of Steel B is relatively pure. The energy dispersive X-ray spectrometer can detect the Cu-containing residual particles in the crack of Steel A. The results of the chemical composition analysis showed that the content of Cu and Sn from Steel A all is one grade higher than Steel B. It should be taken for granted that Cu was detected in the crack of Steel A. However, the content of Sn in both of them was lower than the spectrum detection limit and could not be detected.
4. Results’ Discussion
4.1. Crack Characteristics
According to the observation and analysis of 34 front axles, the distribution of cracks is irregular macroscopically. Most cracks were present on the R-angle position of the low magnification inspection surface. The macroscopic characteristics of the cracks present as thin and straight single strips, and they propagate longitudinally along the front axle. The cross-section features of the cracks are thick in the middle and thin at both ends. The crack tail is tortuous, tapering, and has the intergranular propagated features. Especially, the cracks in the near-surface layer show a “beak-shaped” characteristic with an angle of 45° to the surface. In addition, the cracks present tearing dimple characteristic, indicating they propagate from the middle to both ends.
4.2. Cracking Properties
The main section shows an intergranular fracture and the cross-section cracks’ propagation is intergranular. The section observation shows that there is obvious iron oxide coating in the crack and on both sides of the crack tail, indicating that the cracks have undergone high temperature process and have existed before tempering. The microstructure around the crack is normal tempered sorbite, and there is no oxidation and decarbonization. The above phenomenon indicates that the crack did not exist before quenching heating, i.e., the crack is quenched, longitudinal intergranular crack.
4.3. Quenching Longitudinal Crack Characteristics
In the process of production, longitudinal cracks often occur on completely quenched workpieces. When the component is completely quenched, the internal and external hardness are similar. However, organizational transformation does not occur simultaneously. As the surface gets cold quickly during quenching, the transition from austenite to martensite occurs first. When the surface martensite has been completed, the transformation from austenite to martensite begins in the center. Due to the large specific volume of martensite, the transformation stress finally forms tensile stress at the surface and compressive stress in the core. Longitudinal cracks are formed when the tangential tensile stress on the surface is greater than the axial tensile stress and exceeds the breaking strength of the steel.
The hardenability layer of 42CrMo steel analyzed in this paper is generally about 20 mm, the front axle shape is complex, and the local section thickness is more than 70 mm. In order to ensure the maximum homogeneity of quenching, heat treatment and the quenching process are bound to bring quenching stress, i.e. , the longitudinal cracking of quenching is difficult to avoid.
The tendency of steel to form longitudinal cracks is related to the following factors: (1) Carbon content. With an increase in the carbon content and the increase of solid solution carbon content in martensite, the effect of transformation stress increases and the tendency of longitudinal cracking increases. (2) Metallurgical quality. The inclusions and carbides will be linear or banded along the axial direction in rolling or forging when the content of inclusions and carbides is high. The transverse fracture resistance will be much lower than the axial fracture resistance. Therefore, if the metallurgy quality is poor, when under the same quenching stress and even the tangential stress is slightly less than the axial stress, the workpiece can also form longitudinal cracks because of the tangential tensile stress. (3) The influence of the part shape. The shape of the parts has a complex influence on quenching crack. For round sleeve or hollow thick-walled pipe parts, quenching cracks often occur on the inner hole wall. The R-angle position of the complex part shape is more prone to receive stress concentration. Therefore, longitudinal cracks can occur easily. This point is proved by the large number of cracks in the R-angle position of the low magnification inspection surface in this paper. (4) Quenching temperature. Because of the increasing quenching temperature, the austenite grain grows, the fracture resistance of steel reduces, and the tendency of quenching crack increases. Due to the tangential tensile stress being larger than the axial stress, longitudinal cracks occur. Therefore, the corresponding measures can be taken from the above aspects to reduce the probability of longitudinal crack formation.
4.4. Quenching Intergranular Cracking Mechanism
It is generally believed that there are two reasons for the intergranular cracking of steel. One reason is that there is a thin layer of continuous or discontinuous second phase along the grain boundary. The other is the aggregation of impure atoms such as As, P, Bi, and Sn along the grain boundaries. However, a large number of studies have shown that it is not feasible to explain the intergranular cracking of quenched steel from the viewpoint of impurity atoms’ segregation, especially the experimental fact that the tendency of intergranular cracking increases with an increase in quenching temperature. It is undeniable that these impure atoms may have some segregation at the grain boundary; thus, they can play a promoted role in intergranular cracking. However, it is obviously not the root cause of intergranular cracking of quenched steel. The origin of intergranular cracking is due to martensitic transformation with austenite volume expansion.
Volume expansion occurs when austenite transforms into martensite. The volume expansion of martensitic transformation in numerous grains leads to the total volume change of the quenched steel parts. This phase change volume expansion is not equal in all directions, which must result in intergranular stress. The neighboring grains have different orientations obviously; hence, the strains are different in different directions and the stress must have occurred at the grain boundary, especially with the stress concentration occurring at the triangular grain boundary. The total effect of martensite volume expansion should be the sum of the contributions made by the martensite plates of different sizes in each direction. The nonuniform strain or stress is very large. If the carbon content of martensite is higher, the strength of martensite will be higher. Therefore, the effect of reducing the grain boundary strength and weakening the grain boundary bonding will be more significant, and intergranular cracking will occur under the quenching macro-stress or external force. In summary, it is necessary to consider how to reduce the superposition of phase change stress and take some measures to reduce quenching intergranular cracking.
4.5. Material Comparison
The aggregated inclusions were not observed near or at the end of the front axle cracks of Steel A and Steel B. Low magnification microstructure defects were not observed on the low magnification etching surface, such as white flake, skull patch, air bubble, shrinkage cavity, etc.
The microstructure banded segregation existed in both samples of Steel A and Steel B. The microstructure banded segregation of sample Y-1 of Steel B is obvious, but the crack depth of it is the shallowest (3.2 mm). The microstructure banded segregation of sample Y-3 is not obvious, but the crack depth of it is up to 14 mm. Moreover, the microstructure of samples X-4 and Y-2 without cracking showed obvious banded characteristics, which indicated that the banded segregation of the microstructure was not the main reason for the cracking of the samples.
The aggregated distribution of sulfide inclusions in the matrix of Steel A is higher than B, and more strip MnS inclusions are exposed more intuitively on the low temperature brittle section. However, Steel B is relatively pure. The aggregated distribution of sulfide inclusions in Steel A will promote the occurrence of quenching longitudinal crack.
It is worth noting that the 42CrMo materials of Steel A and Steel B both meet the requirements of GB/T3077-1999 for element content. However, the content of S, N, Cu, and Sn in Steel B (sample Y) is lower than that in Steel A, which indicates that the purity of Steel B (sample Y) is higher than that of Steel A (sample X). The energy dispersive X-ray spectrometer was used to detect some Cu-containing residual particles in the cracks of Steel A. Some data have shown that residual elements such as Cu and Sn mainly come from scrap steel. They have a low melting point and are easy to segregate at the grain boundary to form a low melting point enrichment phase. The crack is easy to occur in their production process and propagate along the austenitic grain boundary.
From the perspective of material comparison, it is natural that Steel A has a higher cracking rate than Steel B. Therefore, it is also an effective way to reduce the cracking probability by putting forward higher quality requirements for raw materials.
5. Conclusions and Recommendations
Typical cracks on the I-shaped and leaf springs of the front axle made by Steel A and Steel B with different micro-alloy composition are analyzed and compared. The cracking rate of Steel A is about 7 times higher than that of Steel B. Chemical composition comparison, low magnification etch test, metallographic, and scanning electron microscope analyses were conducted to identify the failure mechanisms’ differences. The results show that Steel B has a lower content of S, N, Cu, and Sn than Steel A. The microstructures are tempered sorbite, and the cracks are with intergranular fracture characteristics. The aggregated distribution of sulfide inclusions in the matrix of Steel A is higher than that of Steel B. It indicates that the purity distinction of raw steel can have a big difference in the cracking rate of the front axle even under the same manufacturing process.
In the selection of materials, higher quality requirements can be put forward for raw materials, the level of inclusion should be reduced, and harmful elements should be removed as far as possible. Considering that the tendency of quenching crack increases with higher carbon content, the carbon content should be controlled to a lower limit within the range of national standard requirements of 42CrMo material and under the requirements of strength and other performance indicators. The quenching crack property increases with an increase in hardenability. Therefore, the total alloying elements that improve hardenability should be in the lower limit during full quenching. As the shape of the front axle in this paper is complex and the thickness varies greatly, some empirical formulas for the formulation of the above parameters should be adjusted according to the actual production to find out the best heat treatment system. In some aspects, it is also worth paying attention to strictly controlling the fluctuations of instruments in the production process and strengthening the standardization of manual operations for ensuring the stability of the parameters of the heat treatment system.
Data Availability
All data included in this study are available upon request to the corresponding author.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Authors’ Contributions
Huan Xue and Zhifen Wang conceptualized the study. ShiYao Huang and Min Zhu curated the data. Zhifen Wang and Shengnan Liu performed the formal analysis. ShiYao Huang investigated the data. Huan Xue and Yansong Zhang designed the methodology. Huan Xue and Zhifen Wang were responsible for project administration. Huan Xue and Shengnan Liu wrote the paper.