アロイ800の液相拡散接合用インサート金属の開発

数理計画法による接合用インサート金属組成の最適化(第1報)

抄録

This report presents a new alloy-designing technique to develop the insert metal for liquid diffusion bonding. A mathematical programming method was developed by modifying the method in two dimensions proposed by I.P. Schagen into multi dimensions. The main procedure of this method is to obtain the optimum point to maximize some index of an objective performance by calculation from the limited experimental data using interpolation. An interpolating function which yields a curved surface passing on all the specified data points was determined assuming the values of data were normally distributed. If the curved surface is once obtained, its general view will give suggestions for further searches and the optimum point can be efficiently searched. This method has been applied for developing the insert metal for liquid diffusion bonding of Alloy 800. As for the candidate insert metals, alloys which had similar composition to that of Alloy 800, but contained some amounts of silicon, boron and carbon as the depressant elements were chosen. The objective function which was introduced as an index of the performance of insert metals involves the melting point, the strength (hardness) and formability of brittle phases in the bonding interlayer as the evaluating factors. Through calculation based on twenty three experimental date, a composition of Fe-20Cr-30Ni-1.5B-0.75C-4Si which optimized the objective function was determined. Scanning electron microscopic analysis revealed that the joint bonded with the newly developed insert metal (Fe-20Cr-30Ni-1.5B-0.75C-4Si) had a sound microstructure without forming brittle phases in the bonding interlayer. This joint also exhibited good tensile properties, that is, its strength and ductility were comparable to those of the base alloy.