The ability of Ni brazes to join Si3N4 is degraded by dissociation of the ceramic at the bonding temperatures and ingress of Si to form brittle silicides. Attempts to mitigate this problem [1-3] have focused on brazes containing Si, which may not be as susceptible to ingress. Thus, use of the BNi 5 braze, an Ni-19Si-19Cr (at%) alloy with solidus and liquidus temperatures of 1080 and 1135 °C has been suggested, but has also been found to suffer from Si ingress [1]. Therefore, in a recent work the use of BNi 5 has been supplemented by two other approaches: sealing of the ceramic surface with a diffusion barrier such as TiN, and using interactions with the metal workpieces to dilute the Si contents of the joints to below the 10% needed to stabilize silicides [4]. This letter describes studies of two dilution processes: dissolution of the workpiece material in the molten braze to increase the solvent volume, and diffusion of Si out of the braze into the workpieces. The first of these options is difficult to model for a braze joint geometry, but the egress of Si can be analysed using standard diffusion equations. Plates of 2 mm thick Fe-17Cr-12Ni-2Mo AISI 316 steel and of 99.9% pure W were chosen for the study. AISI 316 contains the main constituents of many alloys used in demanding applications where performance could be enhanced by ceramic inserts, and the relatively good match with the thermal expansion coefficient of Si3N 4 has led to the selection of W as a buffer insert material in ceramicmetal components. BNi 5 obtained from GTE Wesgo as 0.080 mm foil was used to braze sandwiches of finely abraded and degreased workpieces by heating in a vacuum of 0.01 Pa at 1150 °C for 15 rain and then holding at 1050 °C for 1-15 h. The condition of the joints was assessed by metallographic examination supplemented by Vickers microhardness (VHN) surveys, scanning electron microscopy (SEM) and electron probe microanalysis (EPMA). These observations provided direct evidence as to whether the time spent at 1150 or at 1050 °C was the more important and hence whether workpiece dissolution or Si egress was the dominant process. The joints between AISI 316 workpieces held at 1050 °C for only 1 h had continuous seams of two hard and frequently cracked phases embedded in a
Read full abstract