IMPORTANT — Note that the carbide disc in Fig. 2 had to be made from much smaller pie-shaped carbide segments since a much larger solid disc of carbide would not have worked (it had been tried) because of the significant difference in COE’s between the carbide and the tool steel used for the die.
The die face was machined flat. The carbide pieces, too, were very flat, and cut into close-fitting pie shapes. The carbide pieces were about 3/8” (10mm) thick and were carefully matched for a good fit, with a close, parallel gap-clearance between each one, so that the molten BFM could flow nicely into and through each joint in the carbide assembly.
These carbide pieces were carefully fitted together on a machined-flat fixturing plate (that would go into the brazing furnace), with a temporary frame around it to keep all the segments closely aligned into a circular disc with tight joints. A thin piece of ceramic cloth had been placed on the flat-fixture surface between the fixture and the carbide-pieces to make sure that the carbide pieces did not get brazed to the fixture!
The BFM (BNi-2) foil we used was available to us in 4-inch wide (100mm) strips, approximately 0.002” (0.05mm) thick. We placed several foil strips side-by-side to cover the full diameter of the carbide circular-disc, and then placed a second layer of that foil on top of that, but with the sheets laid at a 90-degree angle to the first layer, as shown in Fig. 2.
On top of this we then placed an 18-in (45 cm) diameter disc of the pure Nickel-200, approximately 5/64” (2mm) thick, to act as a “shock absorber” for the stresses that would be created between the carbide and tool steel during the high-temp nickel-brazing process.
Then, on top of the thick Nickel-200 layer we placed two more layers of the BNi-2 foil, in the same lay-up manner as we did below the Nickel-200 layer, and finally, carefully laid the tool-steel die on top of that. The die would essentially be then acting as a “dead weight” on top of the vertical assembly, pushing all the layers tightly together. As can be readily seen in Fig. 2, we were essentially nickel-brazing the carbide-disc to the ductile nickel-core from one side, and nickel-brazing the tool steel die to the ductile nickel-core from the other side. We wanted that layer of pure-nickel (Nickel-200) to be thick enough so that the diffusion of the BFM into that layer from each side would be minimal since we wanted most of that pure-nickel sheet to remain un-alloyed and as ductile as possible for its “shock-absorbing” purposes.
We carefully aligned all pieces vertically, and then placed two (2) spring-loaded ring pieces around the base of the stacked assembly, tall enough to cover the brazed joint stack-up and a little of the die itself. The surfaces of that spring-loaded fixture were painted with brazing “stop-off” to prevent sticking of the fixture to the stacked assembly being brazed. The Inconel springs allowed the assembly to grow (expand) outward during heating, but kept constant pressure on the stack, even during cooling as the brazed assembly began to contract (shrink) back to its original size.
Thermocouples were placed into holes already drilled into the open side of the die, so they could read temps inside the die very near to the joint-interface between the die and the Nickel-200 layer. We brazed in a vacuum furnace at about 1950F (1065C), which was a little more than 100F (50C) above the liquidus temp for that BNi-2 BFM, meaning that when that brazing-temp (1950F) was reached, the BFM should indeed have fully flowed! We held the furnace at that temp until all the deeply embedded load-TC’s had reached that brazing temp and stayed there for only a few minutes (5-minutes max), since dwelling at brazing temp too long would have allowed the BFM to diffuse too deeply into the ductile Nickel-200 layer, thereby hardening that layer so much that it couldn’t effectively act as a “shock-absorber”.
We then VERY SLOWLY lowered the temp in the furnace (we did NOT merely turn off the vacuum furnace) since we did not want to cause any significant temp-differentials anywhere in the brazed assembly that might cause the carbides to distort or crack. This is ALWAYS very important when brazing carbides or any other metal combinations in which there are significant differences in the COE’s between the two materials being joined.
The front face of the brazed carbide assembly was carefully inspected after brazing, and the BNi-2 BFM had flowed well into all the thin joints between each of the carbide pie-segments. Following inspection, the surface of the brazed carbide disc was processed to a very smooth finish, holes were put into and through the brazed die assembly, and when placed in service, it performed very well for its intended life.
Brazing can be used to very effectively join materials together that have widely different Coefficients Of Thermal Expansion (COE’s), even when the parts being joined are very large. The design of the joint is very important, and the judicious use of “shock absorber” layers should be considered. Of equal importance is the actual furnace brazing cycle used. It is very, very important that such assemblies be very slowly heated up to brazing temp, using a number of buried TC’s to verify that the temps at different locations inside the assembly are kept quite close together throughout the entire heating and cooling cycles. Otherwise, when large delta-T’s, i.e., temp differentials, occur in such assemblies, it inevitably can lead to distortion or cracking of the assembly. Let me stress again (pun intended) that after the assembly has briefly soaked at brazing temp, it MUST be very slowly cooled back down to room temp to avoid distortion or cracking.
Thus, when the joint design has been properly made, and the brazing cycle is properly controlled, you can have a high degree of confidence that brazing will give you the superb results you want.