The kinetics of hydrogen attack of steels
- PDF / 1,533,954 Bytes
- 15 Pages / 594 x 774 pts Page_size
- 3 Downloads / 254 Views
G. SUNDARARAJAN and PAUL G. SHEWMON are Graduate Student and Professor, respectively in the Metallurgical Engineering Department, Ohio State University, Columbus, OH 43210. Manuscript submitted July 9, 1980. METALLURGICAL TRANSACTIONS A
that the results can be best explained through a grain boundary diffusion controlled bubble growth model, though the correspondence was not good. In case of alloy steels considerable data is available in the literature 7-1~regarding the effects of alloying additions on HA resistance of steels. A number of workers7,8,9have noted a direct correlation between the stability of carbides present in the steel and resistance to HA. This observation can be easily explained without resorting to any particular model since more stable carbides result in lower methane pressure within the bubble on the basis of thermodynamic arguments alone. On the other hand both Naumann, 7 and Rosenthal, e t al, 11 observed that even small additions of Mo (less than 1/2 pct) can increase the incubation period significantly. This observation suggests an interaction between the creep strength of steel and its HA resistance, since Mo increases the creep strength of the steel but does not have a significant effect on the activity of carbon in the steel. 12Finally, neither the grain boundary diffusion model nor the power-law creep model can adequately explain the activation energies and pressure exponents obtained by Sundararajan and Shewmon 5 in case of HSLA steels. Thus the existing HA models provide an inadequate explanation of both the carbon steels and the alloy steel data obtained in the incubation period. One reason for this is that none of the models described earlier are complete. All the existing models tacitly assume that both surface diffusion of Fe atoms along the cavity surface and the processes which accommodate the thickening of the eavitated grain boundary are too rapid to be rate controlling. The model presented below relaxes these assumptions and obtains significantly better agreement with observations. II. THE MODEL The present model deals with the discrete bubble growth stage rather than the fissure stage. If we consider an isolated methane bubble along the grain boundary, its Continued growth, apart from the methane pressure which provides the driving force, presupposes the kinetic steps, shown in Fig. 1. In the sequence shown in
ISSN 0360-2133/81 / 1012-1761500.75/0 9 1981 AMERICAN SOCIETY FOR METALS AND THE METALLURGICAL SOCIETY OF AIME
VOLUME 12A, OCTOBER 1981--1761
Fig. 1, the slowest of sequential steps and fastest of the parallel steps control bubble growth rate. Before discussing the model in detail consider the effect of the extent of grain boundary cavitation on the overall sample expansion rate. Whether the bubble grows by diffusional or power-law creep processes, the final effect is the rigid displacement or the creep of the grains adjoining the bubble at a certain constant rate. At one extreme when all the grain boundaries have bubbles on them, the overall sample expansion rate
Data Loading...
