ICF10B

309L stainless steel overlay and compressive in the ferritic substrate. Within the overlay, the magnitude of the stresses generally increase as the measurement location increases in reference to the notch tip such that the stresses at the notch tip are generally the smallest in magnitude. The residual stresses reach a maximum value of approximately 560 MPa, but are generally on the order of magnitude of 400 MPa or less for this dissimilar metal weld. Within the substrate, the compressive residual stresses reach a maximum value of approximately 130 MPa. Fatigue Crack Propagation Testing Figure 5 shows the C(T) specimens for the 309L stainless steel/A285 carbon steel samples in the AW (5a and 5b) condition following fatigue crack propagation testing. As evidenced from Figure 5a, which corresponds to test identification 309DMW3, the crack grew nearly parallel to and along the fusion line for this specimen. However, Figure 5b, which corresponds to test identification 309DMW2, illustrates that the fatigue crack grew across the fusion line and into the substrate and remained in the substrate during propagation. Figure 6 exhibits the interaction of the crack with the fusion line microstructure for test specimen 309DMW3 (corresponding to Figure 5a), which macroscopically followed the interface. As evidenced from this figure, the extremely hard martensite layer, shown by Figures 1 and 2, does not appear to detrimentally affect the fatigue behavior as might be expected from such a hard layer. That is, propagation does not occur within the martensite layer. The interaction of the fatigue crack with the martensite layer in Figure 6a is representative of only a localized region of the crack growth. Figure 6b, which exhibits the fatigue crack in a different region of the same specimen as in Figure 6a, is more indicative of the entire crack path, such that little interaction exists with the martensite layer. Crack growth rates for the two 309L stainless steel welds, in addition to wrought 309L stainless steel and A285 specimens, are plotted in Figure 7. As evidenced from this figure, the dissimilar metal welds exhibit increased fatigue resistance as compared to the wrought stainless steel and carbon steel specimens. Crack growth into the substrate, which experiences compressive residual stresses, most likely contributes to the enhanced fatigue response. Even the specimen that propagated a small distance through the martensite layer exhibited an improved response, indicating a possible retardation of the fatigue response caused by the martensite layer. Regardless, the main point to emphasize is that, as evidenced by this figure, in conjunction with Figures 5 and 6, the martensite layer does not appear to detrimentally affect the fatigue crack growth behavior of these austenitic/ferritic dissimilar alloy welds. CONCLUSIONS Residual stress measurements made within the vicinity of the fusion line for the dissimilar alloy welds indicated compressive behavior in the substrate and tensile behavior in the overlay. Fatigue crack propagation results indicated that crack growth either crossed over the martensite layer and fusion line and continued within the compressive substrate or continued in the vicinity of the martensite layer within the overlay. Regardless, the hard martensite layer did not negatively affect the fatigue response of the welded materials, as the cracks never propagated fully within the martensite layer. In addition, the crack growth rate curves for the weld overlay specimens exhibited increased fatigue crack propagation resistance as compared to the wrought specimens, most likely due to the compressive residual stresses within the substrate region in which the cracks propagated. Based on these results, the martensite layer does not appear to present a major concern for contributing to interfacial failure in weld overlay applications in austenitic/ferritic dissimilar metal welds. ACKNOWLEDGEMENTS The authors would like to thank Dr. C. Robino at Sandia National Laboratories for the electron beam welding, A. Benscoter for assistance with sample metallurgical preparation, G. Kozma for assistance with the testing equipment, and M. Rex for assistance with specimen machining. This research was sponsored by the American Welding Society Navy Joining Center Fellowship and a consortium of power generation companies. REFERENCES 1. Nelson, T., Lippold, J., and Mills, M. (1998). Science and Technology of Welding and Joining 3, 249-255.

RkJQdWJsaXNoZXIy MjM0NDE=