expected, the failure along the prior austenite grain boundaries should be preferable. The contradiction can be explained on the basis of the following assumptions: a) at a lower hydrogen content, H occupies mainly high-angle boundary sites (i.e. prior austenite boundaries) whereas at higher H contents more hydrogen occupies low-angle boundary sites (martensite laths) or microstructural heterogeneities (carbide precipitates, manganese sulphide, etc.) within grains; b) continuum plasticity theory predicts that even at negligibly small load a maximum stress is attained ahead of the crack tip. For crack propagation the plastic zone size must exceed a critical value, which might be related to characteristic distances such as a prior austenite grain diameter, the mean distance between carbide precipitates, etc [5]. Therefore the observed intergranular fracture for quenched and tempered material is probably a consequence of hydrogen segregation to prior austenite grain boundaries and a resulting lower cohesive force in those regions. In this case the characteristic distance may be the prior austenite grain diameter. For plated materials the hydrogen at martensite laths, carbide precipitates, etc. may play a more important role and the characteristic distance could be related to the spacing of such microstructural heterogeneties. (a) (b) (c) 40 µm 40 µm 40 µm Figure 4: Slow-crack propagation area for quenched and tempered material (a), plated and unannealed (b) and plated and annealed (c) materials. The application of a low temperature annealing on the material was observed to increase the threshold stress intensity factor (cracking started at higher stress level compared to unannealed material Table 3). The fracture surface for annealed samples was revealed to be similar to that for unplated material (Figure 4b). The benefits of annealing in releasing the hydrogen introduced during plating have been investigated by Rebak et al [6]. It was shown that a zinc layer offered a strong barrier to hydrogen escape. In the light of this observation, the increase in resistance to hydrogen induced cracking may instead relate to a redistribution of hydrogen within the steel. According to Townsend [7] hydrogen, introduced during plating, is driven into deep trap sites where it loses mobility and is not available to affect embrittlement. The lower K1H value for plated and annealed material compared to unplated material might be related to the generation of hydrogen at the crack tip via an electrochemical reaction with
RkJQdWJsaXNoZXIy MjM0NDE=