Effects of Silicon Additions and Retained Austenite on Stress Corrosion Cracking in Ultrahigh Strength Steels ROBERT O. RITCHIE, M. H. CASTRO CEDENO, V. F. ZACKAY, AND E. R. PARKER A study has been made of the effects of silicon additions and of retained austenite on the stress-corrosion cracking (SCC) behavior of commercial ultrahigh strength steels (MSI 4340 and 300-M) tested in aqueous solutions. By comparing quenched and tempered structures of 4340 and 300-M i) at equivalent strength and if) at their respective optimum and com- mercially-used heat-treated conditions, the beneficial role of silicon addition on SCC re- sistance is seen in decreased Region II growth rates, with no change in KISCC. The bene- ficial role of retained austenite is demonstrated by comparing isothermally transformed 300-M, containing 12 pct austenite, with conventionally quenched and tempered structures of 300-M and 4340, containing less than 2 pct austenite, at identical yield strength levels. Here, the isothermally transformed structure shows an order of magnitude lower Region II SCC growth rates than quenched and tempered 300-M and nearly two orders of magnitude lower Region II growth rates than 4340, KISCC values remaining largely unchanged. The results are discussed in terms of hydrogen embrittlement mechanisms for SCC in mar- tensitic high strength steels in the light of fire individual roles of hydrogen diffusivity and carbide type. IT has been reported 1'2 that alloying additions of sili- con to high strength steels, such as AISI 4340, can im- prove resistance to hydrogen-assisted cracking and stress-corrosion cracking (SCC) in aqueous solutions. Furthermore, it has been suggested 3'4 that the presence of retained austenite within a tempered martensitic structure in such steels might also lead to improved resistance. The purpose of this paper is to report some recent data on these effects by examining the susceptibility to SCC of AISI 4340 and 300-M commer- cial ultra-high strength steels in distilled water and 3.5 wt pct NaCI solution. EXPERIMENTAL PROCEDURE The materials studied were of aircraft-quality (vac- uum-arc remelted), of composition (in wt pct) shown below: C Mn Cr Ni Mo Si S P V Cu 4340 0.41 0.80 0.79 1,75 0.23 0.26 0,004 0.006 0.06 300-M 0.42 0,76 0.76 1.76 0.41 1.59 0,002 0.007 0.10 - The composition of 300-M is essentially that of 4340 modified with 1.3 pct silicon. The structures investi- gated were compared after conventional quench and tempering and after isothermal transformation heat- treatments, and are summarized in Table I. Ambient temperature mechanical properties are included in Table II. SCC tests were performed in distilled water ROBERT O. RITCHIE is Assistant Professor, Department of Mechan- ical Engineering,MassachusettsInstitute of Technology, Cambridge, MA 02139_ M. H. CASTRO CEDE'NOis Graduate Student and V. F. ZACKAY and E. R. PARKER are Professors, Department of Materials Science and Mineral Engineering,University of California, Berkeley, CA 94720, where Robert O. Ritchie was formerly affiliated. Manuscript submitted June 30, 1977. METALLURGICAL TRANSACTIONS A and in 3.5 wt pct NaC1 aqueous solution, on fatigue pre- cracked 12.7 mm thick, 1-T compact tension specimens, tested under constant load, where growth rates were continuously monitored using a displacement (clip) gage. Values of KISCC, the threshold stress intensity, were defined in terms of the maximum stress intensity at which no crack propagation could be detected within I00 h. The volume percentages of retained austenite in the structures investigated were assessed using standard X-ray and magnetic saturation techniques. Identifica- tion of the principal carbide type was determined from diffraction patterns of extraction replicas of polished and over-etched metallographic specimens. These data are also listed in Table I. RESULTS The variation of rates of SCC (da/dt) as a function of instantaneous stress intensity (K) for all structures, tested in distilled water, is shown in Figs. 1 and 2 and relevant data is listed in Table II. These plots show several clearly distinguishable regions of crack growth which are characteristically observed for environmen- tally-induced cracking; s namely, Region I, where the growth rate is strongly dependent on K, Region II, where the rate is essentially independent of K, and Region III, which is not always observed, where the growth rate increases rapidly as K approaches Kic , the fracture toughness. The influence of silicon additions to 4340 is illus- trated in Fig. 1 where 4340 and 300-M are compared in their respective optimum (and commercially uti- lized) quenched and tempered conditions, namely 4340 tempered at 200~ and 300-M tempered at 300~ De- spite the increase in strength shown by 300-M, Region II growth rates are a factor of four lower than in 4340; the value of KISCC remaining unaffected. A similar com- 1SSN 0360-2133/78/0110-0035500.75/0 91978 AMERICAN SOCIETY FOR METALS AND VOLUME 9A, JANUA RY 1978-3 5 THEMETALLURGICAL SOCIETY OF A1ME