AbstractThe Pd containing brazing alloys were invented as filler materials that (i) had good spreading and gap filling properties, (ii) did not badly erode the high temperature metals for which their use was intended, (iii) were ductile, and (iv) did not contain volatiles, which made their use in radio valves attractive particularly for glass–metal seals, where the Ag–Cu eutectic was a favoured alternative, but could lead to cracking of parent metal during the brazing cycle unless preplated with Ni. The work reported here traced the mechanism of attack of Fe–Ni alloy by Ag–Cu eutectic with three modifications containing nominally 5, 10, and 25 wt-%Pd. A static beam loading technique was used in conjunction with scanning electron and optical microscopy, electron probe microanalysis, X-ray (XPS), and Auger (AES) spectroscopy. The alloys containing 5%Pd or less were hot short, but were ductile when cold bend tested; those containing 10 and 25%Pd were ductile at all temperatures tested. Ductility was associated with a dark phase which concentrated and spread as the Pd content was increased. It appeared that liquid metal embrittlement was responsible for cracking found in the lower Pd content alloys. Analyses using AES and XPS showed that Cu was migrating into the Fe–Ni alloy leading to alloying with Ni and weakness at grain boundaries for the Pd free braze. This was accompanied by the presence of Fe and Ni in the braze, with Ni being predominant. As Pd was increased progressively, the ratio of Fe/Ni changed, with Fe becoming predominant. Metallography showed that the Ag–Cu eutectic braze caused pitting and tunnelling at grain boundaries typical of stress corrosion. Little evidence of this was found in alloys with 5%Pd and it was concluded that general corrosion was increased progressively as a result of an increasing take up of Fe by the Pd, hence removing the susceptibility to intergranular cracking.MST/1117