Kinetics of methane bubble growth in a 1020 steel
- PDF / 1,203,971 Bytes
- 8 Pages / 594 x 774 pts Page_size
- 88 Downloads / 216 Views
I.
INTRODUCTION
STEELS exposed to moderately high temperatures and high pressures of hydrogen may, after an incubation period, undergo irreversible internal damage resulting in marked reduction in tensile strength and ductility. This problem, known as hydrogen attack (HA) is due to nucleation, growth, and subsequent coalescence of methane bubbles primarily on the grain boundaries. 5-s HA, driven by a large internal methane pressure generated by the chemical reaction C + 2H2 --~ CH4, has been a potential problem for the petroleum and synthetic ammonia industries for many years. These industries have constructed the so-called Nelson C u r v e s 9 for safe operation limits. These are based on operating failures and/or years of successful operating experience. Recently the development of coal gasification plants has given a new dimension to this problem emphasizing needs for a better understanding of this phenomenon. Accordingly, a number of studies l~ on the nature and kinetics of HA have been conducted. These have concluded that HA occurs in three stages: (a) an incubation stage in which existing submicron voids grow without affecting the macroscopic properties, (b) a rapid attack stage which is generally characterized by the coalescence of the growing bubbles into fissures along the grain boundaries resulting in a very rapid swelling and affecting mechanical properties, and (c) a saturation or carbon exhaustion stage which is a result of extensive decarburization around the fissures. This leads to lower swelling rates. Service life of critical parts undergoing HA may be characterized by the incubation period since this is the period over which the bubbles grow very slowly and in a reversible manner. It is, therefore, desirable to uncover the underlying bubble growth processes to help extend the service life of the components. To study the bubble growth mechanism, the rate of change of sample volume as a function of temperature and pressure of exposure is normally determined either with the help of a dilatometer or by density measurements. Following such kinetic measurements 1~-~4 a
BINAYAK PANDA is Metallurgist, Touchstone Research Laboratory, 112 14th Street, Wheeling, WV 26003. PAUL SHEWMON is Professor and Chairman, The Ohio State University, Department of Metallurgical Engineering, 116 West 19th Avenue, Columbus, OH 43210. Manuscript submitted December 22, 1982. METALLURGICALTRANSACTIONS A
number of bubble growth models 1~ have been proposed during the last few years. Of these, Shewmon's l~ model for the first time described a complete picture of bubble growth suggesting that under high temperature and low hydrogen pressure exposure the grain boundary diffusion of matrix atoms is the rate controlling step for bubble growth. Subsequent studies by Sagues, et al. 14 and Shih7 have revealed that when bubbles grow larger, their growth is controlled by power law creep rather than grain boundary diffusion. While most of the kinetic measurements 11-~4have reflected the bubble growth kinetics in general, very few, no
Data Loading...
