Influence of precipitation conditions on the tooth root and pitting load carrying capacity of carbonitrided and low pressure carburized gears

In this article, new results on the load carrying capacity of carbonitrided and low pressure carburized gears made out of the case hardening steel 20MnCr5 are presented.

By
Rick Frennea
-
0
2696

Today, case hardening represents the standard heat-treatment process for highly stressed gears. The microstructure in the surface layer of the gears generally consists primarily of martensite with a certain amount of retained austenite. In national and international standards, retained austenite contents of up to 25 to 30 vol.-% are allowed with regard to optimum tooth root and tooth flank load carrying capacity. In addition, precipitations are tolerated only up to certain limits.

In a research project between IWT Bremen and FZG, Technical University of Munich, the application and recommended ranges of the standards were deliberately exceeded in order to determine results on the gear load carrying capacity of material structures characterized by comparatively high contents of retained austenite on the one hand and with different carbide respectively carbonitride precipitations on the other. Different heat-treatment processes were used. A selection of the experimental results is presented in this article, with the focus on the results for the material 20MnCr5. In addition to varying the heat treatment, different blasting conditions were investigated.

The variants with alternative surface layer structures show in some cases significantly increased load carrying capacity numbers compared to the gas carburized reference variant, although the surface hardness is significantly below typical values for case-hardened surface layers. The article evaluates the experimentally determined strength values in the context of the current state of knowledge and shows possible explanations for the increased lifetime of the alternative surface layer structures.

1 Introduction

Today, the vast majority of gears in modern power transmission systems are case hardened in order to ensure a sufficient load carrying capacity during operation. The requirements for case-hardened components and their surface layer structures are defined in national and international standards, such as DIN 3990-5 [1] and ISO 6336-5 [2]. In particular, the limitation of retained austenite to a maximum of 30 vol.-% and defined precipitation states should be mentioned.

Surface layer microstructures with properties deviating from the standard have only rarely been investigated to date. In recent years, several research projects with different objectives have been carried out at Gear Research Center (FZG) of the Technical University of Munich. Several different heat-treatment routes have been applied, primarily to generate surface layers with a higher content of retained austenite. [3, 4]

In a joint research project of Leibniz-IWT in Bremen, Germany, and FZG [5], extensive experimental investigations were carried out on gears with “alternative” surface layers. The material and heat treatment as well as the blasting condition of the test gears were varied. In addition to conventionally gas-carburized reference variants, gas carbonitrided and low pressure carburized test variants were investigated. The project focused on the generation of different precipitation states, with finely distributed and network-like precipitations being investigated.

The focus of this publication is on the results of the tooth root bending and pitting load carrying capacity of the 20MnCr5 case hardened test variants. With regard to both the bending strength in the tooth root and the contact fatigue strength of the tooth flank, a significant increase in service life was determined in some cases. The results of the load carrying capacity tests are supplemented by accompanying investigations in order to be able to classify them according to the current state of knowledge and to derive possible causes for the increased service lives.

2 State of the art

A microstructure consisting of martensite with finely distributed retained austenite is considered as most suitable in the case layer of heavy loaded gears, which is based on extensive investigations and included in standardized requirements and recommendations concerning the microstructure in the surface structure. ISO 6336-5 [2] and DIN 3990-5 [1] specify that the case should consist of fine acicular martensite in combination with a maximum content of retained austenite of 30 vol.-%. For the highest quality grade ME, the retained austenite has to be finely dispersed within the component. The necessary metallographic inspections take place on companion heat-treatment batch test pieces or representative test bars, depending on the achievable quality grade. [2] In DIN 3990-5 [1], a maximum amount of retained austenite of 20 vol.-% is acceptable if it exists in a finely dispersed state. AGMA 923-C22 [6] permits a maximum amount of retained austenite of 30 vol.-% as well, and requires a minimum microhardness of 58 HRC in the area of highest retained austenite content.

In addition to the permissible phase fractions, the permissible precipitation state is specified in ISO 6336-5 [2]. For gears of material grade ML, semi-continuous network-like carbide precipitations at the former austenite grain boundaries are permitted. The specifications for material grade MQ permit carbides distributed in the microstructure, provided they are not longer than 0.02 mm. For material grade ME, finely distributed carbides in the microstructure are tolerable if their length does not exceed 0.01 mm. [2] Compared with ISO 6336-5, the requirements in DIN 3990-5 [1] are less restrictive. No requirements are defined for material quality ML, whereas the specifications for MQ and ME only define that network-like or bone carbides must not be visible under 500x optical magnification. [1] AGMA 923-C22 [6] is in accordance with the requirements of ISO 6336-5, whereby the maximum permissible length of the carbides is not specified for Grade 3 (corresponds to material quality ME of ISO 6336-5). It has to be considered that AGMA 923-C22 [6] is not applicable to carbonitrided parts.

Contrary to what is recommended above, several publications discuss the influence of higher amounts of retained austenite on the fatigue strength. Abudaia et al. [7] found that an initial austenite content of about 60 vol.-% in helical test gears resulted in a higher resistance to pitting. The good performance in the tests can be attributed to a localized stress-induced transformation of retained austenite to martensite at the surface resulting in increased residual compressive stresses and surface hardness. [7] Jeddi and Lieurade [8] mention a redistribution of residual stresses and a supplementary hardening as well, and summarize, that the resulting strength and the stability of the residual stresses is improved when more retained austenite is stress-induced transformed to martensite. Vinokur et al. [9] were able to determine a positive influence of the retained austenite on the contact fatigue strength when the carbon content in the case layer exceeds values greater than 0.8 wt.-%. Razim [10] provides a review of the effect of retained austenite under various loading conditions. In most cases, there is no evidence of a general positive influence of retained austenite on fatigue resistance, which can be confirmed by [11]. Higher amounts of retained austenite result in decreasing strength values if the material is not subjected to a cryogenic treatment after the case hardening process. Only the resistance to pitting is slightly improved by the presence of retained austenite in the case. [10]

In addition to gas carburizing and subsequent hardening, carbonitriding processes can also be used to produce case-hardened microstructures with high retained austenite content, see [12]. Unlike gas carburization, gas carbonitriding is distinguished by the addition of both carbon and nitrogen to the furnace atmosphere [13, 14]. Nitrogen stabilizes austenite at lower temperatures [14]. As a result, a higher proportion of retained austenite can remain in the hardened component [15]. In [16], the load carrying capacity results of two carbonitrided and shot-peened variants are presented. In the experimental investigations, the tooth root bending strength values of the test gears were similar to those of the conventionally gas carburized and case-hardened test gears. The differences in tooth root bending fatigue between the steels 20MnCr5 and 18CrNiMo7-6 are negligible. Rolling contact fatigue was not investigated. [16]

A positive influence of carbonitriding followed by shot peening is described in [17]. The authors investigated the pitting strength in disk-on-disk tests. In [18] and [19], carburized and carbonitrided specimens were compared and the carbonitrided ones showed a better performance in uniaxial fatigue tests. Schurer and Gúntner [4] investigated two variants with carbides on the former austenite grain boundaries, which are also mentioned in [16]. The test gears are treated by a low-pressure carburizing process, followed by oil quenching and tempering for two hours at 180°C. The two variants showed a comparable nominal tooth root bending strength in the pulsator tests, and these values are in the same range as the gas carburized and case-hardened reference variants of 20MnCr5 and 18CrNiMo7-6, respectively.

But there are also hints of negative consequences of an increased content of retained austenite. In rotary bending tests, Prado and Arques determined an optimum amount of retained austenite from 20% to 30% in unnotched carbonitrided specimens, which should be reduced to lower values for notched ones. [20]

3 Aim of the investigations

The aim of the presented investigations is to show the potential of surface layer microstructures, which differs significantly from the recommendations and requirements given in established standards such as ISO 6336-5 [2] and DIN 3990-5 [1]. For some variants, the tooth root bending as well as the pitting load carrying capacity is significantly increased. The determined load carrying capacity values are analyzed with respect to the specific microstructures and conclusions on the underlying mechanisms are drawn. The results could establish alternative heat treatment process routes in the gear industry and help to avoid rejects in the gear manufacturing process.

4 Investigated variants and scope of experiments

This article presents the results of the 20MnCr5 variants, which were investigated in the underlying research project. In total, 12 variants made out of the case hardening steel 20MnCr5 were investigated. The chemical composition of the used material batch was analyzed with optical emission spectrography, see Table 1. Furthermore, the minimum and maximum permitted values for the mass fractions according to DIN EN ISO 683-3 [21] are listed. The measured values are within the tolerance consistently.

Table 1: Chemical composition of the investigated batch of 20MnCr5.

The investigated heat-treatment variants are described in Table 2. The reference variant is gas carburized and conventionally case hardened. The two variants GCN/f1 and GCN/f2 with finely dispersed precipitations are gas carbonitrided with different carbon levels and ammonia mass flows in the furnace atmosphere. The discontinuous precipitation networks are generated by low pressure carburizing (LPC/N1) or by gas carbonitriding (GCN/N2), respectively. As an additional variant LPC/RA, a case layer with almost solely retained austenite was strived. For this purpose, a low-pressure carburizing process was applied. However, this variant was gas quenched.

Table 2: Investigated heat treatment variants (pitting investigations for underlined variants).

The six heat-treatment variants were applied on the test gears. Their geometry data are given in Table 3.

Table 3: Main geometry data of the test gears.

The pulsator gears were tested in different peening conditions. A part of the gears was shot blasted by an impeller process at moderate intensity. These variants are marked with the suffix “-B” in the following. One variant was shot peened in the tooth root area (“-P”) to imply increased residual compressive stresses. The remaining variants stayed in the unpeened condition (“-U”). An overview of the variants is shown in Table 4.

Table 4: Overview of all pulsator variants dependent on their peening condition.

The pitting test gears were all shot blasted after the heat treatment for a sufficient tooth root bending strength and then profile ground on the tooth flanks to ensure suitable gear geometry qualities and surface roughness values for the investigations in the FZG back-to-back test rig.

For determining the tooth root bending load carrying capacity of every pulsator test variant, fatigue tests on constant load stages were performed according to [22]. For some variants, S-N-curves with standard statistical coverage (10-12 test points in the endurance range) were determined. For the other variants, a reduced statistical coverage of 6-8 valid tests in the endurance range was applied.

The pitting investigations were carried out in the area of limited life on two discrete load levels with nominal contact stresses of σH0 = 1,700 N/mm2 and σH0 = 1,800 N/mm2, respectively. The contact fatigue limit in the endurance range was not investigated.

5 Test rigs and test conditions

The tooth root bending strength investigations were carried out on a hydraulic and an electro-magnetic pulsating test rig, see Figure 1 left. The test gear is clamped over four teeth between two jaws so that the sinusoidal force is applied near the end of the singular contact area. At this point of the path of contact, the maximum bending stress in running gears can be obtained. For the determination of the tooth root bending strength, investigations in the endurance range were carried out by applying the stair step- method [23]. The maximum number of load cycles per test run was set to 6〈106 in accordance with FVA guideline 563 I [22].

For the pitting strength investigations, a standardized FZG back-to-back test rig with a center distance of 91.5 mm was used, see Figure 1 right. The test rig is described in detail in DIN ISO 14635-1 [24]. The pinion was mounted on the motor side so that the rotational speed was identical with the motor speed of 3,000 rpm (reversed to the figure). The maximum number of load cycles was 50∙106 for the pinion. The load was applied on the side of the gear, which was driven by the pinion. The test gearset was lubricated with the FVA reference oil FVA 3A (mineral oil, ISO VG 100 with 4% Anglamol 99, a sulphur-phosphorus additive package) at an injection temperature of 60°C (±2°C). The injection flow rate was adjusted to 2 l/min. These test conditions were already used in other research projects, e.g. [3, 4, 25].

Figure 1: Electro-magnetic pulsator test rig [26] (left) and FZG back-to-back test rig with a center distance of 91.5 mm acc. to DIN ISO 14635-1 [24] (right).

6 Results of the investigations

Before presenting the results of the fatigue tests, the properties of the test gears will be described with respect to the microstructure in the case, the hardness, the initial residual stresses, and the content of retained austenite before the test runs.

6.1 Microstructure

The microstructures of the heat-treated and shot-blasted pulsator test gears were analyzed by metallographic inspection. The microsections in the unetched and etched condition were recorded in the unground tooth root section near the 30° tangent of unloaded gear teeth, see Table 5. The corresponding microsections for the variants with different peening conditions are not depicted but showed nearly identical properties considering the microstructure. Some slight differences regarding the content of retained austenite could be observed.

Table 5: Microstructures in the tooth root area near 30° tangent of the shot blasted, unground. pulsator test gears (unetched and nital etched with 2% nitric acid).

The microsection of the gas carburized reference variant Ref-B shows a microstructure consisting of martensite and approximately 20 vol.-% retained austenite. At the surface, a hem of troostite (fine lamellar perlite) can be observed. This microstructure is typical for a conventional case-hardening process.

In comparison to the reference, the variant CGN/f1-B shows a microstructure with a much higher content of retained austenite. The martensite is formed acicularly due to the higher carbon content in the surface layer. Many globular nanoscale carbide and carbonitride precipitations are finely distributed in the case, see corresponding unetched microsection. Directly at the surface, a small amount of non-martensitic microstructural constituents can be found. The variant GCN/f2-B shows a similar microstructure compared to GCN/f1-B but with more discontinuous carbides near the surface. The content of retained austenite is in a comparable range.

In contrast to the two gas carbonitrided variants with dispersed precipitations, the low pressure carburizing caused a microstructure consisting of martensite, retained austenite and massive bone carbides on the former austenite grain boundaries (variant LPC/N1-B). This continuous carbide network reaches a maximum depth of about 100 µm. Due to the vacuum heat treatment, there is no intergranular oxidation at the surface. The variant GCN/N2-B is characterized by a gas carbonitriding process and shows a significantly higher amount of retained austenite in the case layer compared to LPC/N1-B and strong precipitation clusters. Additionally, finely dispersed precipitations can be observed.

The variant LPC/RA-B differs considerably from the five other variants. By a low-pressure carburizing process, a surface layer with an extremely high amount of retained austenite > 80 vol.-% was generated. Only a few martensite needles are embedded in the matrix in the case layer.

Generally, the intended microstructures could be adjusted by the different heat treatment processes and investigated in the fatigue tests.

6.2 Hardness profiles and CHD values

The hardness profiles over the depth of the test gears were measured in the unground tooth root area at the 30° tangent point of contact for the pulsator test gears as well as on the ground tooth flank near the pitch circle for the pitting test gears before the test runs. The investigations were conducted according to the Vickers procedure described in ISO 6507-1 [27] on one test gear for each variant. Both flank sides were investigated, and the average value is stated in the following diagrams.

In Figure 2, the hardness profiles of the unground pulsator test gears for the five unpeened variants are shown. Furthermore, the hardness limit of 550 HV 1 for determining the case-hardening depth (CHD550 HV) acc. to ISO 18203 [28] is indicated. The unpeened reference variant Ref-U shows a typical hardness profile for a conventionally case-hardened gear. The surface hardness is about 720 HV 1, and the maximum hardness is close to the surface. The hardness decreases in the direction of the core and it reaches core hardness values of about 350 HV 1, which are common for the case-hardening steel 20MnCr5. In contrast to the case-hardened reference, the gas carbonitrided variants with finely dispersed precipitations (CGN/f1-U and GCN/f2-U) show lower surface hardness values of 600 HV 1 and 650 HV 1, respectively. The hardness maximum (approximately 750 HV 1) is in a depth of 0.4 mm to 0.6 mm. In greater depths, the hardness decreases in a similar way compared to the case-hardened reference. The low pressure carburized variant LPC/N1-U has a surface hardness > 700 HV 1 and a maximum of 750 HV 1. A significant decreasing hardness toward the surface cannot be observed, but the core hardness is higher than for the other variants. The gas carbonitrided variant GCN/N2-U shows a significant decrease, again, with a surface hardness of 600 HV 1.

Figure 2: Hardness profiles of unpeened pulsator test gears (measured in the unground tooth root at 30° tangent, mean values of left and right flank) and hardness limit of 550 HV 1 acc. to ISO 18203 [28].

Figure 3 depicts the corresponding hardness profiles for the shot-blasted and the shot-peened variants. The case-hardened, shot-blasted reference Ref-B shows a typical hardness profile like the corresponding, unpeened variant. For the two shot-blasted variants with finely distributed precipitations, similar surface values for the unpeened counterparts were investigated. The variants with precipitation networks (LPC/N1-B and GCN/N2-B) show a smaller decrease in hardness toward the surface, each. The hardness profile for the low pressure carburized variant LPC/RA-B with a high amount of retained austenite in the case is clearly different to the other ones. The hardness drops steeper to the core than for the other variants, resulting in a much lower case-hardening depth (CHD), see also Figure 5. There is a remarkable decrease in the surface hardness (about 650 HV 1) and a hardness maximum of approximately 760 HV 1 in a depth of 0.3 mm. The core hardness reaches values of only 310 HV 1 due to gas quenching instead of oil quenching.

Figure 3: Hardness profiles of shot blasted and shot peened pulsator test gears (measured in the unground tooth root at 30° tangent, mean values of left and right flank) and hardness limit of 550 HV 1 acc. to ISO 18203 [28].

For the ground pitting test gears, the hardness depth profiles on the tooth flank near the pitch circle are arranged in Figure 4. The results of the corresponding pulsator variants are mainly confirmed, except for the core hardness of the variant LPC/N1, which is lower compared to the two other variants. For the pulsator test gears, a contrary behavior could be observed.

Figure 4: Hardness profiles of pitting test gears (measured on the ground tooth flank near the pitch circle, mean values of left and right flank) and hardness limit of 550 HV 1 acc. to ISO 18203 [28].

Based on the hardness depth profiles, the corresponding values of case-hardening depth (CHD) are determined. For the 12 investigated pulsator variants, the results can be taken from Figure 5. All variants except LPC/RA exceed the common recommendations for optimal CHD regarding tooth root breakage acc. to [29] and [30] (CHDF,opt ≈ 0.1…0.2 mn). The highest CHD can be observed for the variant GCN/f1 in the unpeened condition.

Figure 5: Case hardening depth (CHD, mean values) of pulsator test gears (measured in the unground tooth root at 30° tangent) and recommendation acc. to Tobie [29] respectively FVA 271 [30] for case carburized materials with lower hardenability.
Figure 6: Case hardening depth (CHD, mean values) of pitting test gears (measured on the ground tooth flank near pitch diameter) and recommendation acc. to ISO 6336-5 [2].

The CHD values for the pitting test gears are shown in Figure 6. All variants are within the recommendations for optimal CHD regarding contact fatigue (CHDH,opt ≈ 0.15…0.4 mn) according to [2]. The case-carbonitrided variant GCN/f1 shows the highest CHD, again.

6.3 Residual stresses and retained austenite

Besides the metallographic investigations, all variants were investigated by X-ray diffraction to determine the initial residual stresses in radial direction and the amount of retained austenite in the case layer after the manufacturing process. The pulsator test gears were analyzed in the tooth root area whereas the pitting test gears were investigated near the pitch circle on the tooth flank. The residual stresses were calculated according to the sin2ψ procedure. Further information on the X-ray measurement parameters is given in Table 6.

Table 6: Main parameters of the X-ray diffraction measurements.

The surface values for the residual stresses are plotted in Figure 7. The standard deviations of each measurement are indicated with error bars. The values are partly higher for the variants LPC/N1 and GCN/N2 due to a more inhomogeneous microstructure. Possible cross-influences from the precipitations should be considered in the interpretation. The unpeened variants have compressive residual stresses in a range from –50…–200 N/mm2, except variant GCN/N2-U, for which a significantly higher scattering has to be mentioned. The shot blasted variants show residual stresses from –350…–530 N/mm2, which is in the lower range for shot-blasted gears. With a controlled shot peening process, the residual stresses can be raised significantly, as it can be seen for variant GCN/f1-P (σRS,surface < –600 N/mm2).

Figure 7: Residual stresses of pulsator test gears (measured by X-ray diffraction in the unground tooth root at 30° tangent, surface values).
Figure 8: Amount of retained austenite of unground pulsator test gears (measured by X-ray diffraction in the tooth root at 30° tangent, median values with minimum and maximum).

The X-ray diffraction measurements enable the determination of the content of retained austenite in the case layer. Figure 8 shows the median values, which are calculated from 5 to 9 measurement points over the depth up to 0.35 mm. The error bars indicate the minimum and maximum values whereby the measurement points close to the surface are excluded because of the influence of the intergranular oxidation zone. It can be stated that the conventionally case-hardened reference shows the lowest values (about 30 vol.-%) of all investigated variants as it was intended. The gas carbonitrided variants with finely dispersed precipitations are characterized by case layers with a content of retained austenite in a range from 60 vol.-% up to 70 vol.-%. In comparison, the variants LPC/N1 with carbide networks have a lower amount of retained austenite, especially the shot blasted variant (~ 40 vol.-%). For the GCN/f2 variants, the content is 60 vol.-% (non-blasted) or 55 vol.-% (shot blasted), respectively. Variant LPC/RA-B shows an amount of about 60 vol.-% retained austenite, but with a maximum of above 80 vol.-%, which is the highest value of all variants.

The variants GCN/f1 and GCN/f2 show no significant change in the amount of retained austenite when shot blasting is applied compared to the unpeened condition. This suggests a stabilization of retained austenite caused by the finely dispersed precipitations. For the variants LPC/N1 and GCN/N2, a slight decrease of the content of retained austenite can be stated after the shot blasting.

The corresponding results for the ground pitting test gears are composed in Figure 9. On the left side, the residual stresses are plotted. It can be seen that the reference variant has the smallest residual stresses in the case. The highest value could be determined for the low pressure carburized variant LPC/N1.

Figure 9: Results of X-ray diffraction measurements on the ground tooth flank of pitting test gears near the pitch circle – left: residual stresses (surface values); right: amount of retained austenite (median values with minimum and maximum).

Variant GCN/f1 is classified in between with a value of about –500 N/mm2. A strong influence of the grinding process on the residual stresses measured directly at the flank surface should be considered in the interpretation of the results.

On the right side of Figure 9, the contents of retained austenite are shown. The reference has a typical value of approximately 25 vol.-% and, therefore, the smallest number. The highest value is reached by variant GCN/f1 with more than 60 vol.-%. The low pressure carburized variant shows a value of nearly 50 vol.-% retained austenite in the case layer.

6.4 Fatigue tests

The tests on the tooth root bending strength were conducted on pulsating test rigs like there were described above. For the 12 variants, S-N-curves were determined but with different statistical coverage.

Figure 10 shows the results of the pulsating tests in the endurance range, presenting the tooth root bending strength for 50% failure probability. The error bars indicate the scattering calculated acc. to Probit [31].

Figure 10: Nominal tooth root bending strength (50% failure probability) for the investigated variants, scattering calculated acc. to Probit procedure [31].

The reference Ref-U shows a σF0∞,50% value of about 920 N/mm2, which corresponds to typical values for case-hardened and non-blasted gears. The corresponding gas carbonitrided variant GCN/f1-U with finely distributed precipitations in the case is increased in its bending strength by about 17%. The other variant with dispersed precipitations, GCN/f2-U, shows a further improvement in the bending strength with σF0∞,50% > 1,200 N/mm2 (+36% compared to the reference). A decreasing bending strength can be stated for the variant LPC/N1-U with carbide networks on the former austenite grain boundaries (σF0∞,50% ≈ 850 N/mm2). This is equivalent to a decrease of 7% compared to the non-blasted reference. The gas carbonitrided variant GCN/N2-U shows more favorable results with a σF0∞,50% value > 1,200 N/mm2.

After applying shot blasting to the test gears, a higher tooth root load carrying capacity can be determined throughout. For the gas-carburized reference, the difference between the non-blasted and the shot-blasted state accounts for 19%. On that basis, the bending strength for the gas-carbonitrided and shot-blasted variant GCN/f1-B can be increased on a similar scale like in the unpeened condition. By the application of a shot-peening process, the bending strength can be increased even more significantly.

The variant GCN/f2-B has a σF0∞,50% of about 1,300 N/mm2, which is 20% above the shot blasted reference Ref-B. The low pressure carburized and shot blasted variant LPC/N1-B shows a higher tooth root load carrying capacity in comparison to the case-carburized reference despite the carbide networks in the case. The highest tooth root load carrying capacity could be determined for the variant GCN/N2-B.

The variant with a high amount of retained austenite (LPC/RA-B) has a bending strength of σF0∞,50% ≈ 1,300 N/mm2, although the CHD is considerably lower than for the other investigated variants.

For all variants, an improvement of the tooth root bending strength by the application of a shot blasting or shot peening process is evident. The increase can be quantified in a range of 5% up to 41%. The effect is most pronounced for the variant LPC/N1. The qualitative influence of precipitations is discussed in more detail below.

The investigations regarding pitting took place in the range of limited life. Three variants were selected based on the tooth root bending results. In Figure 11, the number of load cycles until pitting failure is plotted with respect to the two load levels (σH0 = 1,700 N/mm2 and σH0 = 1,800 N/mm2, respectively). On each load level, three test runs were conducted and the average lifetime for 50% failure probability was calculated based on a Weibull distribution [32].

Figure 11: Results of pitting strength investigations for the investigated variants (average failure load cycles calculated with Weibull distribution).

The case-carburized reference shows average failure load cycles of about 17〈106 (σH0 = 1,700 N/mm2) and 11·106 (σH0 = 1,800 N/mm2), respectively. The other two variants demonstrate a potential increase of the pitting lifetime of nearly 50% on both load levels each.

6.5 Damage patterns

Following the fatigue tests, the damage patterns of the gears were investigated by optical microscopy. Table 7 gives an overview of exemplary fracture surfaces in the endurance range for all six heat treatment variants as determined in the pulsator tests for tooth root bending fatigue strength. Only the shot blasted variants are depicted. For all variants except the low pressure carburized variant LPC/RA-B, similar damage patterns could be observed. The damage pattern of LPC/RA-B is more rugged compared to the other variants, which could be explained by the lower core hardness.

Table 7: Exemplary damage patterns of shot blasted pulsator tests gears in the endurance range (crack initiation on bottom in each picture).
Table 8: Exemplary damage patterns on pinion of pitting tests at a nominal contact stress of σH0 = 1,700 N/mm2.

Typical pitting damages on tooth flanks of pinions after the test run are compiled in Table 8. The Ref variant is characterized by a severe pitting in the flank area with negative specific sliding up to the pitch circle. Near the tooth root fillet at the start of active profile, micropitting can be observed. The flank of the variant GCN/f1 shows a single pitting in the center of the flank area with negative specific sliding.

Furthermore, some micropitting is present. The damage pattern of the variant LPC/N1 shows a higher area with micropitting, which is grown along the face edge up to the pitch circle. A smaller pitting is located in the area of negative specific sliding starting in the micropitted area.

7 Discussion

7.1 Tooth root bending strength

Systematic investigations on the tooth root load carrying capacity were conducted to determine the potential of gears with alternative case layer microstructures compared to a typically case-hardened reference microstructure. The results of the pulsating tests were presented above and are classified in the state of knowledge in the following.

The nominal stress numbers for bending σF lim were calculated from the nominal tooth root load carrying capacity acc. to ISO 6336-3 [33], respectively. The results are inserted in the diagram for case hardened wrought steels acc. to ISO 6336-5 [2] (see Figure 12). For comparison, a reference value for conventionally case-hardened gears of 16MnCr5 (comparable to 20MnCr5), taken from [34], is plotted.

Figure 12: Experimentally determined nominal stress numbers for tooth root bending, plotted in the diagram for case hardened wrought steels acc. to ISO 6336-5 [2]; 16MnCr5 reference taken from [34].

The high load carrying capacity numbers for bending of the variants with finely distributed precipitations in the case layer can be explained by the two hardening mechanisms applied to the test gears. By the quenching during the heat-treatment process, the austenite is transformed to martensite predominantly, which causes a significant increase in hardness and strength. By the precipitation of carbides or carbonitrides due to higher contents of carbon and nitrogen, a second hardening mechanism is applied. The finely dispersed, nanoscale precipitations cause an impediment of the movement of dislocations in the material matrix. Therefore, the local strength around the fine precipitations rises and in sum, the global strength of the whole component can be improved. The application of shot blasting or shot peening after the thermochemical heat treatment leads to another increase in strength as it is known for conventionally gas carburized, case hardened gears.

For the test gears with network-like carbides in the case (LPC/N1-U and LPC/N1-B), a slight decrease of the tooth root load carrying capacity could be determined in the unpeened condition (minus-7% compared to the corresponding reference). This is due to local notch effects caused by the accumulation of brittle carbides on the former austenite grain boundaries, which facilitate the crack propagation. In the blasted condition, this negative effect of the presence of semi-continuous carbide precipitations is superimposed by the contribution of the compressive residual stresses in the case layer.

The variants GCN/N2 show strength values comparable to the other gas carbonitrided variants GCN/f1 and GCN/f2. This can be explained by the microstructure of these variants. Beside network-like precipitations, finely dispersed precipitations could be found in the case layer. Therefore, two competing effects affect the resulting tooth root load carrying capacity. The influence of the fine, nanoscale precipitations is dominant toward the massive semi-continuous carbides/carbonitrides.

The high bending strength of the low pressure carburized variant with a high amount of retained austenite can be repatriated to the comparatively high residual stresses. Among all shot blasted variants, LPC/RA-B has the highest compressive residual stresses near the surface (σRS ≈ –530 N/mm2).

The retained austenite ensures a certain ductility of the microstructure, so the crack initiation and propagation are inhibited or decelerated, respectively. In addition, stress peaks can be reduced by the stress-assisted transformation from retained austenite to martensite under load.

The results prove microstructures differing from the requirements and recommendations of the standards such as e.g. ISO 6336 [2] and DIN 3990 [1], offer the potential for increased tooth root bending strength. Therefore, the microstructures have to be analyzed in more detail with respect to their stability and possible changes under load.

7.2 Pitting load carrying capacity

The results of the contact fatigue investigations indicate a significantly positive influence of case layer structures with higher contents of retained austenite on the pitting load carrying capacity. Compared to the case-carburized reference, lifetime increase potentials up to about 50% could be determined. The main reasons are the residual stresses primarily caused by the grinding of the tooth flanks and the amount of retained austenite. The case layer structures with higher contents of retained austenite retard the crack propagation due to the higher ductility. The hardness and the CHD show no direct correlation to the pitting strength and play a subordinate role within the herein performed investigations. From the pictures of the damage patterns, the conclusion can be drawn that the variant LPC/N1 with carbide networks is more exposed to micropitting than the other two variants.

8 Conclusion and outlook

In this article, new results on the load carrying capacity of carbonitrided and low pressure carburized gears made out of the case hardening steel 20MnCr5 were presented. The tooth root bending strength of 12 variants was investigated with respect to different heat-treatment routes and peening conditions. For the pitting strength, three selected variants were investigated in screening tests in the regime of limited life.

The results were evaluated in terms of the achieved material properties and microstructures.

The results demonstrate the potential of case layer microstructures differing from the common requirements and recommendations according to international and national standards, such as e.g. ISO 6336 [2] and DIN 3990 [1]. Common case-carburized microstructures are well established at the current time, but the presented results indicate possible positive effects by increased amounts of retained austenite in the case layer. Especially shot-blasted variants with finely distributed precipitations in the case layer reach higher tooth root bending strength numbers compared to a corresponding, conventionally case-hardened reference. A general negative influence of network-like carbides in the case on the load carrying capacity can be avoided if a shot-blasting process is applied after the heat treatment. Regarding the pitting strength, lifetime could be increased by nearly 50% compared to the reference due to the alternative case layer microstructures. Possible complications during gear grinding due to the higher amount of retained austenite should be considered.

In pursuing investigations, the precipitations in the case layer should be characterized in more detail, e.g. regarding their size, morphology, and quantitative distribution within the matrix. Furthermore, the stability of the individual phases and structural components has to be analyzed to prevent improper changes in the microstructure and early failures of components during operation.

Acknowledgement

The presented results are based on the research project IGF no. 20054 N undertaken by the Research Association for Drive Technology envy. (FVA) supported by the German Federation of Industrial Research Associations e.V. (AiF) in the framework of the Industrial Collective Research Programme (IGF) by the Federal Ministry for Economic Affairs and Climate Action (BMWK) based on a decision taken by the German Bundestag. The authors would like to thank the sponsorship and support received from the FVA, AiF, and the members of the project committee. 

Bibliography

  1. DIN 3990-5: Calculation of load capacity of cylindrical gears – endurance limits and material qualities (in German: Tragfähigkeitsberechnung von Stirnrädern – Dauerfestigkeitswerte und Werkstoffqualitäten), Deutsches Institut für Normung e.V., Berlin, 1987.
  2. ISO 6336-5: Calculation of load capacity of spur and helical gears: Part 5: Strength and quality of materials, International Organization for Standardization, 2016.
  3. S. Lombardo, M. Steinbacher, T. Tobie, B. Clausen, B.-R. Höhn, and H.-W. Zoch, FVA No. 513 I – Issue 970 – Carbozahn – Final report: Carbonitriding of gears (in German: Carbonitrieren von verzahnten Getriebebauteilen). Frankfurt a.M., 2011.
  4. S. Schurer et al., FVA No. 513 III – Issue 1248 – Randschichtgefüge – Final report: Alternative multi- phase surface layer structures in case hardening for increasing the strength properties of gears (in German: Alternative mehrphasige Randschichtgefüge beim Einsatzhärten zur Steigerung der Festigkeitseigenschaften von verzahnten Getriebebauteilen). Frankfurt a.M., 2017.
  5. N. Blech, H. Surm, T. Tobie, M. Steinbacher, K. Stahl, and R. Fechte-Heinen, FVA No. 513 IV – Issue 1541 – CarbidRAnd – Final report: Influence of nanoscale precipitation states in the surface layer by thermochemical heat treatments on the gear load carrying capacity (in German: Einfluss von nanoskaligen Ausscheidungszuständen in der Randschicht durch thermochemische Wärmebehandlungen auf die Zahnradtragfähigkeit). Frankfurt a.M., 2023.
  6. Metallurgical Specifications for Steel and Cast Iron Gearing, 923-C22, American Gear Manufacturers Association, Alexandria, United States, 2022.
  7. F. Abudaia, J. T. Evans, and B. Shaw, “Characterization of Retained Austenite in Case Carburized Gears and Its Influence on Fatigue Performance,” ASM Proceedings: Heat Treating, Volume 1, pp. 62–69, 2000.
  8. D. Jeddi and H.-P. Lieurade, “Effect of retained austenite on high cycle fatigue behavior of carburized 14NiCr11 steel,” Procedia Engineering, no. 2, pp. 1927–1936, 2010, doi: 10.1016/j.proeng.2010.03.207.
  9. B. B. Vinokur, S. E. Kondratyuk, L. I. Markovskaya, R. A. Khrunik, A. A. Gurmaza, and V. B. Vainerman, “Effect of Retained Austenite on the Contact Fatigue Strength of Carburized Steel,” Met Sci Heat Treat, 1978, doi: 10.1007/BF00713758.
  10. C. Razim, “Retained austenite – on the state of art about the cause and effects in case-hardened steels (in German: Restaustenit – zum Kenntnisstand über die Ursache und Auswirkungen bei einsatzgehärteten Stählen),” Härterei-Technische Mitteilungen, vol. 40, no. 4, pp. 150–165, 1985, doi: 10.1515/htm-1985-400402.
  11. W. Beumelburg, “Behaviour of case-hardened specimens with different surface conditions and surface carbon contents in rotating bending, static bending and impact bending tests (in German: Das Verhalten von einsatzgehärteten Proben mit verschiedenen Oberflächenzuständen und Randkohlenstoffgehalten im Umlaufbiege-, statischen Biege- und Schlagbiegeversuch),” Dissertation, University of Karlsruhe (TH), Karlsruhe, 1973.
  12. M. Steinbacher, F. Hoffmann, H.-W. Zoch, S. Lombardo, and T. Tobie, “New case layer structures of carbonitrided components and their strength properties – Part 1: Investigation of material properties (in German: Neue Randschichtgefüge carbonitrierter Bauteile und deren Festigkeitseigenschaften – Teil 1: Untersuchungen werkstofftechnischer Eigenschaften),” HTM Journal of Heat Treatment and Materials, 2015, doi: 10.3139/105.110266.
  13. DIN EN ISO 4885: Ferrous materials – Heat treatments – Vocabulary (in German: Eisenwerkstoffe – Wärmebehandlung – Begriffe), International Organization for Standardization, Berlin, 2018.
  14. D. Liedtke, “Technical bulletin 452 – Case hardening (in German: Merkblatt 452 – Einsatzhärten),” Wirtschaftsvereinigung Stahl, Düsseldorf, 2008.
  15. C. Gu, B. Lou, X. Jing, and F. Shen, “Mechanical Properties of Carburized Cr-Ni-Mo Steels with Added Case Nitrogen,” J. Heat Treat., vol. 7, no. 2, pp. 87–94, 1989, doi: 10.1007/BF02833193.
  16. K. J. Winkler, T. Tobie, K. Stahl, C. Güntner, and S. Schurer, “Material Properties and Tooth Root Bending Strength of Shot Blasted, Case Carburized Gears with Alternative Microstructures,” in AGMA Fall Technical Meeting 2019, Detroit, 2019.
  17. A. Goloborodko and Y. Watanabe, “Effect of shot peening after carbonitriding on the contact fatigue strength of chromium-containing steel,” 2008.
  18. C. Kanchanomai and W. Limtrakan, “Effect of Residual Stress on Fatigue Failure of Carbonitrided Low-Carbon Steel,” Journal of Materials Engineering and Performance, vol. 17, no. 6, pp. 879–887, 2008, doi: 10.1007/s11665-008-9212-x.
  19. C. Stöberl, H. Bomas, B. Clausen, F. Hoffmann, and H.-W. Zoch, “Investigations on the endurance limit of carbonitrided specimens of case hardening steels 18CrNiMo7-6 and 20MnCr5 (in German: Untersuchungen zur Dauerfestigkeit carbonitrierter Proben aus den Einsatzstählen 18CrNiMo7-6 und 20MnCr5),” Materialwissenschaft und Werkstofftechnik, vol. 46, no. 6, pp. 533–549, 2015, doi: 10.1002/mawe.201500272.
  20. J. M. Prado and J. L. Arques, “Influence of retained austenite on the fatigue endurance of carbonitrided steels,” Journal of Materials Science, no. 19, pp. 2980–2988, 1984, doi: 10.1007/BF01026976.
  21. DIN EN ISO 683-3: Heat-treatable steels, alloy steels and free-cutting steels – Part 3: Case- hardening steels, International Organization for Standardization, Berlin, 2019.
  22. Forschungsvereinigung Antriebstechnik e.V., FVA guideline No. 563 I: Standardisation of load carrying capacity tests – Recommendations for the standardisation of load carrying capacity tests on quenched and hardened cylindrical gears (in German: Vereinheitlichung von Tragfähigkeitsversuchen – Empfehlungen zur Vereinheitlichung von Tragfähigkeitsversuchen an vergüteten und gehärteten Zylinderrädern). Frankfurt a.M., 2012.
  23. M. Hück, “An improved procedure for the evaluation of stair step tests (in German: Ein verbessertes Verfahren für die Auswertung von Treppenstufenversuchen),” Materialwissenschaft und Werkstofftechnik, vol. 14, no. 12, pp. 406–417, 1983, doi: 10.1002/mawe.19830141207.
  24. DIN ISO 14635-1: Gears – FZG test procedures – Part 1: FZG test method A/8,3/90 for relative scuffing load-carrying capacity of oils (in German: Zahnräder – FZG-Prüfverfahren – Teil 1: FZG- Prüfverfahren A/8,3/90 zur Bestimmung der relativen Fresstragfähigkeit von Schmierölen), Deutsches Institut für Normung e.V., Berlin, 2006.
  25. N. Blech, C. Güntner, S. Schurer, M. Steinbacher, T. Tobie, and K. Stahl, “Material influence on mechanical properties of gears with alternative microstructures,” Forschung im Ingenieurwesen/Engineering Research, 2023, doi: 10.1007/s10010-022-00608-9.
  26. C. Weber, T. Tobie, and K. Stahl, “Rapid and Precise Manufacturing of Special Involute Gears for Prototype Testing,” in AGMA Fall Technical Meeting 2019, Detroit, 2019.
  27. ISO 6507-1: Metallic materials – Vickers hardness test: Part 1: Test method, International Organization for Standardization, 2018.
  28. ISO 18203: Steel – Determination of the thickness of surface-hardened layers, International Organization for Standardization, 2016.
  29. T. Tobie, “On the pitting and tooth root load carrying capacity of case-hardened gears – Influences from case hardening depth, heat treatment and manufacturing with regard to the size (in German: Zur Grübchen- und Zahnfußtragfähigkeit einsatzgehärteter Zahnräder – Einflüsse aus Einsatzhärtungstiefe, Wärmebehandlung und Fertigung bei unterschiedlicher Baugröße),” Dissertation, Technical University of Munich, Munich, 2001.
  30. T. Tobie, P. Oster, and B.-R. Höhn, FVA No. 271 – Issue 622 – Härtetiefe-Großzahnräder – Final report: Influence of case hardening depth on the pitting and tooth root load carrying capacity of large gears (in German: Einfluss der Einsatzhärtetiefe auf die Grübchen- und Zahnfußtragfähigkeit großer Zahnräder). Frankfurt a.M., 2001.
  31. T. Hösel and F. Joachim, “Tooth flank contact strength taking into account the probability of failure (in German: Zahnflankenwälzfestigkeit unter Berücksichtigung der Ausfallwahrscheinlichkeit),” Antriebstechnik, vol. 17, no. 12, pp. 533–537, 1978.
  32. M. Hein, M. Geitner, T. Tobie, K. Stahl, and B. Pinnekamp, “Reliability of Gears: Determination of Statistically Validated Material Strength Numbers,” in AGMA Fall Technical Meeting 2018, Detroit, 2018, pp. 1–22.
  33. ISO 6336-3: Calculation of load capacity of spur and helical gears: Part 3: Calculation of tooth bending strength, International Organization for Standardization, 2019.
  34. G. Niemann and H. Winter, Machine elements (in German: Maschinenelemente – Band 2: Getriebe allgemein, Zahnradgetriebe – Grundlagen, Stirnradgetriebe), 2nd ed. Berlin: Springer-Verlag, 2003.

Printed with permission of the copyright holder, the American Gear Manufacturers Association, 1001 N. Fairfax Street, Suite 500, Alexandria, Virginia 22314. Statements presented in this paper are those of the authors and may not represent the position or opinion of the American Gear Manufacturers Association. (AGMA) This paper was presented Octobber 2023 at the AGMA Fall Technical Meeting. 23FTM05