Influence of High Electrical Currents on WEC Formation in Rolling Bearings

Premature bearing failures due to white etching cracks (WECs) occur when the rolling contact is subjected to a so-called additional load such as an electrical current flow through the bearing in addition to the pure rolling load. Tests with oil-lubricated roller bearings and greased ball bearings showed that high electrical currents are a very critical WEC trigger. The influence of current intensity, electrical polarity, and load type (stationary ring load or rotating ring load) on WEC formation was proven by the performed bearing tests, whereas the influence of lubricant was relatively low. Based on the findings, the failure hypothesis "energetic WEC fatigue" was extended for high electrical currents.

Keywords: White etching cracks; WEC; white structure flaking; brittle flaking; hydrogen; electric current; rolling bearing

Introduction

White etching cracks (WECs), also known as brittle flaking, irregular white etching areas, or white structure flaking, are a subsurface bearing fatigue failure. Rings and rolling elements of all rolling bearing types ([1], [2]) independent of the type of lubrication (oiled or greased lubricated) ([3]) and heat treatment (through- or case-hardened) ([4], [5]) can be affected.

WECs are characterized by mostly irregular networks of cracks at and in white etching areas (WEAs), which do not have the form of white bands or butterflies. The WEAs consist primarily of nanocrystalline ferrite with dissolved carbides that appear white if polished and etched with Nital ([[6]]). Three forms of WEC appearance can be observed (see Fig. 1). Figure 1 Microcuts in which the WEA network dominates and cracks seem to be only flanking are most frequently shown in literature, though sometimes the opposite occurs and cracks are then in a majority. Furthermore, in some cases long oriented WEAs can be found. They seem to be degenerated butterflies because the length of a butterfly is basically only 10 to 250 µm (wing tip to wing tip) ([6], [13]).

PHOTO (COLOR): Figure 1. Types of rolling contact fatigue without WEA formation (subsurface-initiated fatigue, surface-initiated fatigue) and with WEA formation (white bands, butterflies, WEC). There are three forms of WEC appearance (WEA network flanked by cracks, WEA structure consisting of butterflies, WEC network flanked by WEAs).

The root cause of WECs is not fully understood. Tests with hydrogen precharged bearings proved that an increased hydrogen content can initiate WEC failures ([10], [[14]]). Moreover, a hydrogen weakening of the bearing steel at the beginning of the operating life is sufficient to trigger WECs ([11]). Several researchers ([[17]]) have shown that atomic hydrogen is generated by high rolling contact friction or electric currents leading to an increased hydrogen concentration in the bearing components. The hydrogen generation due to certain types of corrosion, especially in the presence of water and a direct current voltage ([22]), is also well-known. Therefore, the "hydrogen WEC path" shown in Figure 2 gives a possible explanation for why WECs are often observed in applications with high friction energy ([23]), corrosion risk ([24]), or current passage ([3]).

PHOTO (COLOR): Figure 2. State-of-the-art and subjects of research regarding WEC formation.

Currently, researchers are debating the metal physical processes behind crack and WEA formation (stress-induced carbon diffusion, localized plastic deformation, recrystallization, cracking). In this context, whether the crack ([14]) or the WEA ([25]) forms first is discussed. Based on general material science work, both scenarios are possible in the case of hydrogen-assisted WEC formation. Hydrogen promotes cracking (hydrogen-enhanced decohesion, hydrogen-induced cracking) as well as plastic deformation (hydrogen-enhanced local plasticity). Moreover, there are other theories, including a crack rubbing mechanism ([5]) as the cause of WEAs, where hydrogen can enhance this rubbing mechanism.

Perhaps the dominating process depends on the operating conditions (the magnitude and relation of hydrogen content to the Hertzian stresses). This would explain the different WEC appearances shown in Fig. 1.

Moreover, some researchers work on an alternative WEC path, where an increased hydrogen content is not necessarily required ([12], [[26]]). For example, Gould et al. ([28]) showed that WEC formation can occur due to an electrical current despite using a hydrogen-free lubricant with minimum water content. Accordingly, Gould et al. ([28]) concluded that local plasticity possibly enhanced by electric current could have aided crack formation.

One source of bearing currents can be electrostatic charges produced by the principle of a Van de Graaff generator. This occurs, for example, between the rollers of paper machines or film stretching lines ([29]), in belt drives, and also within the rotors of wind turbines ([30]). These typically very small direct currents, often less than 100 µA, usually do not lead to any surface damage. However, they can cause early WEC failures ([3], [24]).

Converters are another source of electrical bearing currents. Using a converter to generate variable frequencies (speeds) in electrical drives leads to parasitic effects like additional power losses or electrical bearing currents, which are mainly caused by the common mode voltage as arithmetic mean of the conductor–ground voltages. This high-frequency voltage generates bearing currents due to capacitive and inductive couplings in the motor ([[31]]). Such relatively high and strongly transient bearing currents can cause premature bearing damage in electrical machines. In addition to typical traces of current passage at the surface like rehardening zones and smoothed melting beads, WEC failures can occur (see Figure 3). Bearings in electric machines as well as bearings in adjoining units like gearboxes can be partly affected.

PHOTO (COLOR): Figure 3. Pittings, traces of current passage, and WECs of an electric motor bearing: 1, cage and ball set from e-motor bearing; 2, pitting on ball; 3, position microcut ball; 4 and 5, WECs in the origin of the pitting; 6, traces of current passage (smoothed melting beads) on the surface of the outer ring; 7, rehardening zone of up to 2 µm beneath the outer ring raceway; 8, WEC network beneath the outer ring raceway.

Current limits are seldom published in the literature. Schenk ([34]) reported 1.4 A/mm2 as the apparent bearing current density (ratio between maximum amperage and the sum of the Hertzian contact area of all rolling elements in contact) at which, after a run time of 500 h, fluting (ripple) leading to unacceptably high running noises may be visible. In view of a run time in the region of 50,000 h, Pittroff ([32]) reduced this limit to 0.7 A/mm2, and when considering field experience in the rail industry, this limit is further reduced to 0.1 A/mm2. Below this value, classical current damage due to passage of electrical current is not expected.

Published test results ([35]) indicate that these high electrical currents can also lead to WEC failures. This was observed in tests on greased 6203 bearings by Kawamura and Mikami ([36]) where electrical currents (>0.1 A) triggered WEC failures below 100 h. However, in addition to the electrical load, these bearings were subjected to a high angular acceleration, making an isolation of current as the cause of failure impossible. But Oezel ([25]) and Dinter ([35]) were able to show that high bearing current densities (>0.1 A/mm2) are a WEC trigger in rolling contact fatigue tests using a four-disc tribometer. WECs occurred only if either the test specimens were precharged with hydrogen or a high electrical current was applied. Similar results were performed by Gould et al. ([28]) on a three-ring-on-roller test (micropitting rig). Starting at a current density of approximate 0.2 A/mm2 with simultaneous 30% slip and a relative film thickness Λ0 of about 0.5, WECs were formed after less than 300 million load cycles. The influence of the lubricant on the run time was approximately a factor of 2. Guzman et al. ([37]) were able to confirm these findings on cylindrical roller bearings. Inner rings and rollers failed due to WECs only if the inner ring was connected anodically (i.e., connected to the positive pole of the current source). Contrary to these findings, only cathodically connected rings (i.e., connected to the negative pole of the current source) were affected by WECs if small currents were applied ([3]). This shows that the relationship between electrical current and WEC failure are poorly understood so far.

To better understand the influence of current intensity, electrical polarity, and load type (stationary ring load or rotating ring load) on WEC formation, the following bearing tests with oil-lubricated roller bearings and greased ball bearings are discussed.

Test conditions

The influence of high electrical currents was tested both on oil-lubricated roller bearings and greased ball bearings (100Cr6, martensite). Oil-lubricated NU207 cylindrical roller bearings (d = 35 mm, D = 72 mm, B = 17 mm) were run on the electrically modified R4NN test rig (Fig. 4). Here, four rolling bearings were loaded identically and lubricated by a synthetic industrial gearbox oil ISO VG 320, hereinafter referred to as I1. A speed of 4,500 rpm and a outer ring temperature of about 80 °C resulted in consistently good lubrication conditions (Λ0 > 4).

PHOTO (COLOR): Figure 4. R4NN test rig (electrically modified).

The tests with the greased ball bearings were conducted on the idler pulley test rig (BSRA test rig), shown in Figure 5. The test bearing was a modified basic type 6203 ball bearing (d = 17 mm, D = 40 mm, B = 14 mm) with increased load capacity. Three polyurea greases (grease 1: ester base oil, ν40°C = 100 mm2/s; grease 2: ester/ether base oil, ν40°C = 102 mm2/s; grease 3: ester base oil, ν40°C = 160 mm2/s) were used, hereafter referred to as PH/E1, PH/EE2, and PH/E3. All tests were performed at 10,500 rpm leading to good lubrication conditions (Λ0 > 1.5) despite bearing temperatures of about 90 °C and up to 130 °C. The ball bearing was electrically connected via two carbon brushes to the turning outer ring (see Fig. 6). To reliably measure the current through the ball bearing, the inner ring was insulated while mounted to the test rig. Therefore, the current produced by the electrostatic charge of the belt drive could also be measured.

PHOTO (COLOR): Figure 5. Testing principle of BSRA test rig.

PHOTO (COLOR): Figure 6. Equivalent circuit of BSRA test rig with carbon brushes contact.

The specific test conditions for each tested bearing in the current work are shown in Table 1. The electrical voltage drop across the bearing was less than 5 V and mostly less than 3 V. All specified current values are per bearing.

Table 1. Testing parameters for alle conducted tests.

After the test, the parts of the bearings were inspected visually. On parts affected with pittings, a subsurface origin of the damage was confirmed. Finally, microcuts in the circumferential direction through the pitting were performed and etched with 3% Nital (HNO3 in alcohol). On parts without any macroscopic damage, the circumferential microcuts were performed in the load zone of the ring in the axial center of the raceway.

Electrical properties of a rolling bearing

The electrical properties of a rolling bearing are determined by the elastohydrodynamic rolling contacts. Figure 7 shows a sufficient equivalent circuit of a rolling bearing.

PHOTO (COLOR): Figure 7. Equivalent circuit diagram of a rolling bearing.

The separating lubricating film acts not only as an electrical capacity but also as an ohmic resistance, which is very high. In the field of full lubrication it lies mostly in megohm range due to the low specific conductivity of typical lubricants. In the mixed friction area, the ohmic resistance breaks down by many powers of 10 due to the metallic contact of the surface roughness.

However, at very small currents (<100 µA) the field intensity in the lubricating gap already exceeds the disruptive strength of the lubricant and the boundary layers. At a lubricating thickness of 0.1 µm, typical for the ball bearing used, a voltage drop of 6.5 V across the lubricating gap leads to an electrical field of 65 kV/mm. This value corresponds to typical disruptive strength of oils. The magnitude of the discharge resistance is fluctuating in time and is typically far less than 10 Ω. Once there were disruptive discharges through the lubricating film, which could be detected by an oscilloscope, the mean electrical resistance of the rolling bearing decreased with the amperage, whereas the discharge frequency increased (see typical current–voltage characteristic for the tested bearings in Fig. 8).

PHOTO (COLOR): Figure 8. Typical current–voltage characteristic for the tested ball bearings and direct current.

Failure behavior of oil-lubricated roller bearings

An apparent bearing current density of 0.01 A/mm2, well below the limit values given in the literature for classic current damages (>0.1 A/mm2), was sufficient to generate WECs in cylindrical roller bearings (see Figs. 9 and 10). Traces of current passage (electro-erosive wear, smoothed melting beads) were clearly visible on the raceway surfaces. Moreover, the current led to premature bearing failures. The achieved WEC run times were significantly lower than the calculated rating life Lhmr according to ISO/TS 16281 (38) (Lhmr = 2,030 h). Although the maximum Hertzian pressure on the outer ring (1,900 N/mm2) was significantly lower than that on the inner ring (2,400 N/mm2), only the outer rings were affected. Therefore, the stationary ring load showed a higher WEC risk than the rotating ring load.

PHOTO (COLOR): Figure 9. WEC failures due to high electricalcurrents (NU207, oil: I1, pHz,max, OR = 1,900 N/mm2; pHz,max, IR = 2,400 N/mm2, J = 0.01 A/mm2).

PHOTO (COLOR): Figure 10. Pitting, traces of current passage, and WECs below the outer ring surface of a tested cylindrical roller bearing (582 h, NU207, oil: I1, pHz,max, OR = 1,900 N/mm2, J = 0.01 A/mm2).

The parasitic bearing currents in electric drives are predominantly alternating currents, which can also lead to WECs, as Figure 11 shows. As in the tests with direct current, the outer rings failed. However, differences were observed in the direct current tests depending on whether the outer ring was connected to the positive pole or negative pole of the current source. When the outer ring was the anode, the outer ring tended to fail a little bit earlier than in the tests with the alternating current. When the outer ring was the cathode, only a single WEC failure occurred after a relatively long run time.

PHOTO (COLOR): Figure 11. WEC failures due to high electrical current only at the outer ring (NU207, oil: I1, pHz,max, OR = 1,900 N/mm2; pHz,max, IR = 2,400 N/mm2, J = 0.01 A/mm2).

Failure behavior of greased ball bearings

Greased ball bearings are also quickly harmed by high electrical currents. In addition to WEC-caused pitting formations, fluting (ripple) occurred in many of the tested bearings. Figure 12 shows typical fluting patterns on the inner rings. These periodically recurring areas of high current mark density lead to high running noises of the bearings and make it difficult to detect WEC-related pittings. Consequently, some bearings failed without WEC-caused pittings (see suspended parts in Figure 13, Figure 14 with a short run time).

PHOTO (COLOR): Figure 12. Fluting (ripple) due to high electrical current on inner ring raceways (6203mod,grease: PH/E3, pHz,max = 2,150 N/mm2, J = 0.4–1 A/mm².

PHOTO (COLOR): Figure 13. Influence of current intensity on the WEC run time (6203mod, PH/E1, pHz,max, IR = 2,150 N/mm2).

PHOTO (COLOR): Figure 14. Influence of current intensity on the WEC run time (6203mod, PH/EE2, pHz,max, IR = 2,150 N/mm2).

Although the rating life Lhmr ([38]) of the tested bearings was relatively long (>10,000 h), the bearings failed very early due to the passage of current. As seen in Figure 15 WECs occurred after less than 100 h.

PHOTO (COLOR): Figure 15. Influence of current direction on WEC formation (6203mod, 1 A/mm2, PH/EE2, pHz,max, IR = 2,150 N/mm2, pHz,max, OR = 2,000 N/mm2).

The WEC life and the affected ring depended on the current direction. The combination rotating outer ring and constant load direction due to the belt preload caused a higher load cycle frequency of the inner ring than the outer ring. Moreover, the maximum Hertzian pressure at the inner ring (2,150 N/mm2) was a little bit higher than that at the outer ring (2,000 N/mm_SP_2_sp_). Nevertheless, WECs occurred in the outer ring if the current direction was positive (i.e., the outer ring acted as the anode). Therefore, the anodically connected rings were significantly more WEC-critical. The effect seemed to be even stronger than for the cylindrical roller bearings.

Figure 16 shows typical WEC networks in the failed rings that were found in polished and etched microcuts beneath the surface, both near and away from the pitting. Moreover, a thin white layer near the surface was mostly visible. These rehardening zones are typical indications of high bearing currents, which are characterized by many flash discharges causing very high local temperatures for a very short time. Based on the very low electrical resistance during a flash discharge, the average voltage drop across the roller bearing, despite the high amperage, was less than 3 V in almost all tests and therefore significantly smaller than in the comparable bearing test with only 40 µA (up to 65 µA) as described by Loos et al. ([3]).

PHOTO (COLOR): Figure 16. WEC network beneath the surface and near the pitting plus rehardening zones at the surface of an inner ring in an etched microcut (83 h, 1 A, PH/EE2, pHz,max, IR = 2,150 N/mm2).

As expected, the WEC life depends strongly on the amperage. The greater the current, the earlier the failure. This behavior applies to all tested greases (see Figs. 13, 14, 17) as well as positive (see Figs. 13 and 14) and negative current direction (see Fig. 17). The higher WEC criticality of the anode is also evident here. The influence of the grease seemed to be comparatively low. The WEC life of grease PH/E1 was about a factor of 1.5 greater than that of grease PH/EE2.

PHOTO (COLOR): Figure 17. Influence of current intensity on the WEC run time (6203mod, PH/E3, pHz,max, OR = 2,000 N/mm2).

Rolling bearings of electrical motors in industrial application are often subjected to low loads. This leads to high apparent current densities due to a small total Hertzian contact area for the same amperage. Single tests indicate that under such operating conditions fluting failures are more likely to occur than WEC failures (see Figure 18). Three out of four bearings failed with fluting before 500 h despite low bearing load. The run times of bearings with cathodic inner rings tended to be shorter than those with anodic rings. WECs did not form in all of these cases.

PHOTO (COLOR): Figure 18. Failure tendency at low load (6203mod, PH/EE2, direct current, pHz,max, IR = 1,300 N/mm2).

Discussion of results

Passage of a high electrical current (I > 10 mA, J > 0.001 A/mm2) can lead to early WEC fatigue damage in roller bearings as well as in ball bearings. Similar to the tests with small currents ([3]) (I < 1 mA, J < 0.0001 A/mm2) and with high friction ([2]), the bearing rings subjected to a stationary ring load showed a higher WEC tendency than those subjected to a rotating ring load. In a cylindrical roller bearing the outer ring always failed with WECs due to the passage of current, although the maximum Hertzian pressure of the inner ring was much greater than that of the outer ring. This is contrary to the fatigue mechanism on which the rating life calculation ([38]) is based.

The hydrogen root cause path described in Fig. 2 provides a possible explanation for this behavior. With a rotating ring load the time for hydrogen uptake (overrolling of material element) in relation to the time of hydrogen emission (between two overrollings) is shorter. This leads to a lower mean hydrogen concentration on the surface (see also ([3])). Moreover, in the case of stationary ring load this stronger hydrogen-weakened material element is more often subjected to a fatigue-relevant load in the middle of the load zone.

In contrast to the WECs created under electrostatic load, characterized by small direct currents (<0.1 mA, J < 10−4 A/mm2), the anodically charged ring is far more WEC-critical at high currents. This fits the findings of Guzman et al. ([37]) and Kawamura and Mikami ([36]) for oil-lubricated cylindrical roller bearings and greased ball bearings with relatively high amperages. In these cases bearing rings only failed due to WECs if they were connected anodically. The higher WEC-criticality of the anode at high apparent current densities (0.37 A/mm2) was also confirmed in four-disc WEC tests at a maximum Hertzian pressure of 1,900 N/mm2 (35).

Furthermore, the crucial parameter for the WECs driven by a high electric current is different. The decrease in WEC life under increased amperage observed by Kawamura and Mikami ([36]), Guzman et al. ([37]), and Dinter ([35]) was also detected in the presented tests. The mean voltage drop across the bearing and thus the average electrical field strength across the lubricating gap was relatively low. In the case of WEC formation due to electrostatic load the opposite occurred: The voltage drop and field strength in the elastohydrodynamic lubrication contacts were high and correlated with the WEC life ([3]). Here, the amperage was very small and not a suitable WEC indicator ([3]).

Fluting is a bearing failure mechanism that runs parallel to WEC formation at high currents. Low-loaded bearings failed only due to fluting, whereas in moderate- and high-loaded bearings the WEC formation mostly limited the run time. This confirms the WEC formation hypothesis according to Fig. 2. In addition to WEC triggers like an electrical current, a fatigue-relevant load is needed for WEC formation.

Root cause hypothesis

The distinction between two WEC mechanisms (cathodic WEC fatigue vs. energetic WEC fatigue), as shown in Figure 19 (39), provides a possible explanation for the described contrary behavior at high and low currents. The cause–effect relationships for both WEC mechanisms are also summarized in Fig. 19.

PHOTO (COLOR): Figure 19. Root cause hypothesis for WEC formation due to bearing currents.

The cathodic WEC fatigue triggered by a constant direct current only occurs on the bearing ring connected to the negative pole of the current source and therefore acts as a cathode. This and the strong influence of the lubricant formulation, usually greater than a factor of 10 (40), indicates that electrochemical processes similar to the cathodical stress corrosion cracking are involved in the WEC formation.

The reason for cathodical stress corrosion cracking defects is the hydrogen absorption of steels at the cathode ([41]), which strongly decreases the strength. Such hydrogen absorption can also lead to WEC failures as shown by Vegter and Slycke ([15]), Ruellan et al. ([42]), Hamada and Matsubara ([43]), and Szost and Rivera-Díaz-del-Castillo ([44]). Average hydrogen concentrations of less than 1 ppm are sufficient to trigger WECs ([11]). This concentration of hydrogen might be below the detection line ([45]).

A prerequisite for the hydrogen absorption for cathodic WEC fatigue is the existence of solvated hydrogen cations (protons) and therefore electrochemical reactions where they are formed. Only at the cathode does the Volmer reaction take place, which leads to hydrogen adsorption on the steel surface. Some of this atomic hydrogen reacts to molecular hydrogen (Tafel reaction), and a small part is absorbed by the steel surface and diffuses into the material ([15]).

The kinetics of several electrochemical reactions can mostly be well explained by the Butler-Volmer equation. The electrochemical current and thereby the material conversion per time are, among others, dependent on the voltage as soon as this exceeds a threshold. However, this equation is only valid for high diffusion rates with high ion mobility. In lubricated elastohydrodynamic lubrication contacts the ion mobility should be considered small, so that it is likely that the migration velocity of the ions determines the electrical current. Assuming that the forces on the ions based on the concentration differences can be neglected in comparison to the forces based on the very high electrical fields (>5 kV/mm) in the lubricating gap, the electrochemical current, or rather the hydrogen deposition rate, is proportional to the electrical field (Nernst-Planck equation). This provides an explanation as to why the electrical field intensity correlates to the WEC tendency and does not require a flash discharge in the lubricating film ([3]).

The energetic WEC fatigue is triggered by high electrical currents. In converter-fed machines, such high and very transient bearing currents can occur as a parasitic effect ([31]). The common mode voltage (ucm) and its high gradient (ducm/dt) at the motor terminals cause three types of harmful bearing currents (Electric Discharge Machining) (EDM) currents, rotor ground currents, and circulating currents). The ohmic resistance of a bearing in full lubrication is very high, typically greater than 1,000,000 Ω. Thereby, high currents must flow almost completely in the form of arc discharges, which have a multiphase generation process ([46]). A very high localized field strength at roughness peaks leads to an increase in the oil temperature and conductivity as well as an imprint of preferred directions (percolation). Streamers, thermo-ionized channels with small density, are formed preferentially at the anode ([46]). Therefore, thermal stress of the lubricant and the degree of possible dissociation processes are greater at the anode. If a streamer reaches the counterelectrode, this channel is used for the main discharge connected with vaporizing of the oil, light emission, and voltage breakdown. The flash temperatures are so high that the bearing steel is melted locally and radicals are generated by thermal dissociation of the lubricant (e.g., hydrocarbon or water molecules) or the surface layers (e.g., hydroxides). A small portion of the formed hydrogen radicals can be absorbed at the anodic and cathodic rings and increase the hydrogen content of the bearing components ([20]). Mikami and Kawamura ([47]) showed with axial ball bearing tests under high current passage that electro-erosive wear occurred mainly on the anodic ring. Consequently, the anodic ring has no (or a thin) oxidic tribolayer as a hydrogen absorption barrier. Therefore, the hydrogen concentration and the WEC risk tend to be higher at the anode for energetic WEC fatigue.

Conclusions

The increased use of modern converters in variable-speed drives leads to an elevated risk of current passage in motor and transmission bearings. The maximum amperages of those converter-induced currents are typically relatively high. Tests carried out with roller and ball bearings proved that such high electrical currents may cause early WEC failures as an additional mode to the small electrical currents as previously investigated ([3]).

The following knowledge was gained from the WEC tests with high electrical currents: High direct or alternating currents can lead to premature bearing failure due to WEC formation in oil-lubricated roller bearings as well as in greased ball bearings.

The bearing rings connected to the positive pole of the current source (anode) tend to fail much earlier.

The WEC life decreased with increasing amperage.

The bearing rings subjected to a stationary ring load showed a higher WEC risk than those subjected to a rotating ring load.

Low-loaded bearings tended to fail due to fluting

Based on these findings, and with respect to the current knowledge concerning WEC formation, the failure hypothesis "energetic WEC fatigue" was suggested for high electrical currents:

A high electrical current causes a high frequency of flash discharges in the lubrication gap. Thereby, the electrical charges flow almost completely through the discharge channel in a very short time. The flash temperatures in the narrow channel are so high that the bearing steel is melted locally and radicals are generated by thermal dissociation of the lubricant and the surface layers protecting against adhesive wear. A small portion of the formed hydrogen radicals can be absorbed and increase the hydrogen content of the bearing components. The anodic ring is more WEC-prone because the lubricant near the anode heats up on average more during the arc discharge process and the anode is covered with fewer boundary layers acting as a hydrogen barrier due to electro-erosive wear processes.

NOMENCLATURE

• J

• Apparent current density
• L hmr
• Modified reference rating life
• p Hz
• Maximum Hertzian pressure

• R

• Electric resistance

• U

• Voltage

• U Bearing
• Voltage across the rolling bearing
• u cm
• Common mode voltage (arithmetic mean of the line-to-ground voltages)

• ϑ

• Temperature

• κ

• Viscosity ratio
• Λ0
• Relative film thickness (central film thickness related to sum roughness)
• υ40
• Dynamic viscosity at 40 °C

1 Review led by B. Gould. Ruellan, A., Stadler, K. et al. (2020), "The Influence of Lubricant Formulation on Early Thrust and Radial Bearing Damage Associated with White Etching Cracks," Proceedings of the Institution of Mechanical Engineers - Part J: Journal of Engineering Tribology, 235, 1 – 13. 2 Loos, J., Kruhoeffer, W., Merk, D. et al. (2017), Berechnungsansätze für die WEC-Neigung unter reibenergetischer Wälzbeanspruchung [Calculation Approaches for the WEC Risk under Frictional Rolling Stress], VDI Tagung Gleit- und Wälzlagerungen, Schweinfurt, Germany. 3 Loos, J., Bergmann, I., and Goss, M. (2016), "Influence of Currents from Electrostatic Charges on WEC Formation in Rolling Bearings," Tribology Transactions, 59, pp 865–875. doi: 10.1080/10402004.2015.1118582 4 Blass, T. et al. (2017), " Influence of Material and Heat Treatment on the Formation of WECs on Test Rig FE8," Advances in Steel Technologies for Rolling Bearings STP 1580. 5 Kadin, Y. and Sherif, M. Y. (2017), "Energy Dissipation at Rubbing Crack Faces in Rolling Contact Fatigue as the Mechanism of White Etching Area Formation," International Journal of Fatigue, 96, pp 114 – 126. doi: 10.1016/j.ijfatigue.2016.11.006 6 Rumpf, V. (2018), A Study on Microstructural Alterations in White Etching Cracks, Dark Etching Region, and White Etching Bands in Rolling Contacts, Doctoral Thesis, Southampton, Great Britain. 7 Holweger, W. et al. (2015), "White Etching Crack Root Cause Investigations, White Etching Crack Root Cause Investigations," Tribology Transactions, 58 (1), pp 59 – 69. doi: 10.1080/10402004.2014.942938 8 Diederichs, A. M., Schwedt, A., Mayer, J., and Dreifert, T. (2016), "Electron Microscopy Analysis of Structural Changes within White Etching Areas," Materials Science and Technology, 32, pp 1683–1693. doi: 10.1080/02670836.2016.1139030 9 Paladugu, M. and Hyde, R. S. (2017), "White Etching Matter Promoted by Intergranular Embrittlement," Scripta Materialia, 130, pp 219 – 222. doi: 10.1016/j.scriptamat.2016.11.030 Oezel, M. et al. (2018), "Formation of White Etching Areas in SAE 52100 Bearing Steel under Rolling Contact Fatigue—Influence of Diffusible Hydrogen," Wear, 414–415, pp 352 – 365. doi: 10.1016/j.wear.2018.08.022 Evans, M.-H., Richardson, A. D., Wang, L., and Wood, R. J. K. (2013), "Effect of Hydrogen on Butterfly and White Etching Crack (WEC) Formation under Rolling Contact Fatigue (RCF)," Wear, 306 (1–2), pp 226 – 241. doi: 10.1016/j.wear.2013.03.008 Bruce, T., Long, H., Slatter, T., and Dwyer-Joyce, R. S. (2016), "Formation of White Etching Cracks at Manganese Sulfide (MnS) Inclusions in Bearing Steel Due to Hammering Impact Loading," Wind Energy, 19, pp 1903 – 1915. doi: 10.1002/we.1958 Bhadeshia, H. K. D. H. (2012), "Steels for Bearings," Progress in Materials Science, 57, pp 268 – 435. doi: 10.1016/j.pmatsci.2011.06.002 Ooi, S. W., Gola, A., Vegter, R. H. et al. (2017), "Evolution of White-Etching Cracks and Associated Microstructural Alterations during Bearing Tests," Materials Science and Technology, 33, pp 1657–1666. doi: 10.1080/02670836.2017.1310431 Vegter, R. H. and Slycke, J. T. (2010), "The Role of Hydrogen on Rolling Contact Fatigue Response of Rolling Element Bearings," Journal of ASTM International, 7 (2), pp 1–12. doi: 10.1520/JAI102543 Evans, M.-H. et al. (2013), "Effectof Hydrogen on Butterfly and White Etching Crack (WEC) Formation under Rolling Contact Fatigue (RCF)," Wear, 306 (1–2), pp 226 – 241. doi: 10.1016/j.wear.2013.03.008 Richardson, A. D. et al. (2018), "Thermal Desorption Analysis of Hydrogen in Non-Hydrogen-Charged Rolling Contact Fatigue-Tested 100Cr6 Steel," Tribology Letters, 66, pp 4. doi: 10.1007/s11249-017-0947-0 Kohara, M. et al. (2007), "Study on Mechanism of Hydrogen Generation from Lubricants," Tribology Transactions, 49 (1), pp 53 – 60. doi: 10.1080/05698190500486324 Han, B. et al. (2016), "In Situ Detection of Hydrogen Uptake from Lubricated Rubbing Contacts," Tribology Online, 11, pp 450 – 454. doi: 10.2474/trol.11.450 Gegner, J. et al. (2018), Wälzlagerschäden mit weiß anätzenden Rissen abseits der Windenergie, 3. VDI-Fachkonferenz Schadensmechanismen an Lagern, Aachen, Germany Tamada, K. and Tanaka, H. (1996), "Occurrence of Brittle Flaking on Bearings Used for Automotive Electrical Instruments and Auxiliary Devices," Wear, 199, pp 245 – 252. doi: 10.1016/0043-1648(96)06990-6 Rossmeisl, J. et al. (2005), "Electrolysis of Water on Oxidized Metal Surfaces," Chemical Physics, 319, pp 178 – 184. doi: 10.1016/j.chemphys.2005.05.038 Gould, B. and Greco, A. (2016), "Investigating the Process of White Etching Crack Initiation in Bearing Steel," Tribology Letters, 62, pp 26. doi: 10.1007/s11249-016-0673-z Iso, K., Yokouchi, A., and Takemura, H. (2005), " Research Work of Clarifying the Mechanism of White Structure Flaking an Extending the Life of Bearings," SAE Technical Paper 2005-01-1868. doi: 10.4271/2005-01-1868 Oezel, M. O. (2018), Entstehung von White Etching Areas und White Etching Cracks als Folge der Wälzbeanspruchung im Stahl 100Cr6 [Formation of White Etching Areas and White Etching Cracks as a Result of Rolling Stress in 100Cr6 Steel], Doctoral Thesis, RWTH Aachen, Germany. Scepanskis, M. et al. (2015), "The Numerical Model of Electrothermal Deformations of Carbides in Bearing Steel as Possible Cause of White Etching Cracks Initiation," Tribology Letters, 59, pp 1 – 10. Stadler, K. et al. (2017), Short Term Heavy Loads—One Cause for Premature Bearing Failures and White Etching Cracks, GfT Tagung : Göttingen, Germany. Gould, B. et al. (2021), "The Effect of Electrical Current on Premature Failures and Microstructural Degradation in Bearing Steel," International Journal of Fatigue, 145, pp 106078. doi: 10.1016/j.ijfatigue.2020.106078 Stache, H. and Petri, S. (2008), Protect against Damages Due to Discharge—Electrical Conductive Bearing Grease as Cheap and Effective Solution, Klüber : Munich, Germany. Mendez, Y. et al. (2014), Elektrostatische Aufladung von Rotorblättern und deren Effekte auf das Überspannungsschutzsystem in Windparks [Electrostatic Charging of Rotor Blades and Its Effects on the Surge Protection System in Wind Farms], ETG Fachtagung Diagnostik elektrischer Betriebsmittel, Berlin. Muetze, A. (2003), Bearing Currents in Inverter-Fed AC-Motors, Doctoral Thesis, TU Darmstadt, Germany. Pittroff, H. (1968), " Wälzlager im elektrischen Stromkreis" [Bearings in the Electrical Circuit], Elektrische Bahnen, 3, pp 54 – 61. Tischmacher, H., Gattermann, S. et al. (2010), " Bearing Wear Caused by Converted-Induced Bearing Currents," IECON 2010—36th Annual Conference on IEEE Industrial Electronics Society. Schenk, O. (1953), " Stromdurchgang durch Wälzlager" [Current Passage through Bearings], Der Maschinenschaden, 11/12, pp 131 – 135. Dinter, R. (2018), " Influence of Mechanical and Electrical Loading on the Formation of White Etching Cracks (WEC)," Bearing World Conference, TU Kaiserslautern, Germany, March 6–7, 2018, Forschungsvereinigung Antriebstechnik (FVA). Kawamura, T. and Mikami, H. (2007), "Development of NA103A Long-Life Grease for Automotive Components," NTN Technical Review, 75 (11). Guzman, G., Oezel, M. O. et al. (2017), " Bildung von 'White Etching Cracks' (WEC) durch elektrischen Strom an ölgeschmierten Wälzkontakten" [Formation of White Etching Cracks in Oil-Lubricated Rolling Contacts Due to Electric Current]. Drive Technology Colloquium ATK 2017: Aachen, March 07–08, RWTH Aachen University, ISBN 3-7431-4897-8, 978-3-7431-4897-0, pp 177–194. ISO/TS 16281. (2008), " Rolling Bearings—Methods for Calculating the Modified Reference Rating Life for Universally Loaded Bearings." Loos, J. et al. (2016), "Influences on Generation of White Etching Crack Networks in Rolling Bearings," Journal of Mechanical Engineering and Automation, 6, pp 85 – 94. Franke, J., Loos, J., Kruhoeffer, W., Merk, D., and Blass, T. (2020), " Influence of Oil Formulation on White Etching Crack Formation Depending on WEC Main Mechanism," 22nd International Colloquium Tribology, Esslingen, Germany. 28–30, January 2020, Stuttgart/Ostfildern, Germany. Juilfs, G. G. (2000), Das Diffusionsverhalten von Wasserstoff in einem Niedriglegierten Stahl unter Berücksichtigung des Verformungsgrades und der Deckschichtbildung in Alkalischen Medien [The Diffusion Behavior of Hydrogen in a Low-Alloy Steel Considering the Degree of Deformation and Surface-Layer Formation in Alkaline Medium], Doctoral Thesis, TU Hamburg–Harburg, Germany. Ruellan, A., Ville, F. et al. (2013), " Rolling Contact Fatigue in Rolling Element Bearings for Wind Turbine: A Focus on White Etching Cracks," STLE 68th Annual Meeting & Exhibition, Detroit, MI, May 5 – 9. Hamada, H. and Matsubara, Y. (2006), " The Influence of Hydrogen on Tension-Compression and Rolling Contact Fatigue Properties of Bearing Steel," NTN Technical Review No. 74. Szost, R. H. and Rivera-Díaz-del-Castillo, P. E. J. (2013), "Unveiling the Nature of Hydrogen Embrittlement in Bearing Steels Employing a New Technique," Scripta Materialia, 68, pp 467 – 470. doi: 10.1016/j.scriptamat.2012.11.018 Suter, T. et al. (2009), Lokale Wasserstoffbestimmung auf (hochfestem) Stahl [Local Hydrogen Determination in High-Strength Steel], Institut für Werkstofftechnologie AG : Wallisellen, Switzerland. Kuechler, A. (2018), High Voltage Engineering, Heidelberg, Dordrecht, London, New York : Springer. Mikami, H. and Kawamura, T. (2007), "Influence of Electrical Current on Bearing Flaking Life," SAE Technical Paper 2007-01-0113.

By Joerg Loos; Iris Bergmann and Matthias Goss

Reported by Author; Author; Author