In situ detection of hydrogen evolution during lubricated sliding contact

Rolling contact fatigue (RCF) of bearings is typically manifested by subsurface cracking and eventually spalling [1]. In the case of oil lubricated contact, the damage processes involved in crack initiation and/or propagation may be accelerated in the presence of absorbed hydrogen. This is often associated with the formation of ""white etching areas"" (WEAs), so-called for their appearance in optical micrographs after etching with nital. Bearing failure due to formation of WEA's is commonly referred to as ""brittle flaking"". Brittle flaking represents a serious industrial problem affecting bearings of all sizes in a diverse range of applications. Despite a considerable body of literature, the mechanism for the generation and absorption of hydrogen is still unclear. Numerous investigations were carried out in the 1960's and 1970's with the aim of evaluating water contamination of the lubricant as a possible source of hydrogen. Cantley [2] attributed the correlation between the reduction in bearing lifetime and increasing water content of the lubricant to the influence of hydrogen. Grunberg et al. [3] and Ciruna et al. [4] evaluated the influence of water contamination on fatigue lifetime by using a four ball testing apparatus and measuring the hydrogen concentration of the balls after the tests. Grunberg et al. [3] concluded that hydrogen plays an important role on the lifetime of the rolling elements and that the presence of water in the lubricant reinforces this effect. Ciruna et al. [4] used secondary rotary bomb oxidation tests to demonstrate that the hydrogen is produced in an oxidation reaction of the lubricant with the steel surface. Additives like water or glacial acetic acid influence the amount of hydrogen that is formed [4]. The generation of hydrogen as a product of some tribo-chemical reaction of the lubricant, in the absence of water or other additives, is now a common theme in the literature. Tamada et al. [5] found that brittle flaking occurs in double-roller RCF tests of specimens pre-charged with hydrogen. Based on these results, the authors proposed that brittle flaking was due to hydrogen produced by the reaction of the lubricant with the metal surface and absorbed by the steel body. Kino et al. [6] proposed that the critical factor affecting hydrogen generation is the local increase in temperature due to frictional heating. Various authors [7, 8, 9, 10] have evaluated the mechanisms for hydrogen generation by performing ball-on-plate tribometer tests in a vacuum chamber and measuring the evolved hydrogen using a mass spectrometer; however, vacuum tribometer tests are limited to the evaluation of oils that are stable under vacuum. Based on such tests, Kohara et al. [7] proposed the lubricant decomposition reaction is catalysed by the film-free steel surface, which is exposed by sliding contact. Lu et al. [8] correlated the sliding speed, load and quantity of gaseous reaction products, which include H2+, CH3+, C2H3+, C2H4+, C3H7+, for vacuum tribometer tests using the lubricant Pennzane 2001A. The authors rationalized the results in terms of removal of the oxide layer from the steel surface by rubbing. In two separate papers [9, 10] the same authors showed that the addition of 1 wt% of di-tert-dodecyl disulfide to Pennzane 2001A prolonged the period of time before onset of decomposition. This delay was attributed to the coverage of the steel surface with iron sulfide and hydrocarbon sulfide, which poisoned the catalytic reaction and hindered the decomposition process.