Phase equilibria in the Ni-Si-Mg system and fracture toughness of selected in-situ intermetallic composites

Abstract

The phase equilibria in the Ni-Si-Mg system and mechanical properties such as fracture toughness and yield strength of selected ternary and binary composites in the system have been studied.

The Ni-Si-Mg ternary phase diagram has been established after homogenization and slow cooling to room temperature. The isothermal section of the phase diagram at 900°C in the Nirich region was also established after isothermal annealing at the temperature followed by water quenching. The chemical compositions of the alloys and their phases were obtained using fully quantitative energy dispersive x-ray spectroscopy (EDS) with standard spectrum files created from intermetallic compounds MgiNi and NhSi. The following intermetallic phases have been observed: (a) four new ternary intennetallic phases, designated as v, co, μ, and -c, (b) a ternary intermediate phase, Mg(Ni,Si)2 based on the binary MgNh phase containing Si, (c) three ternary intermetallic phases, T), K, and~, previously reported by the present authors [96Son, 98Son1], and (d) MgiSiNi3 (Fe2Tb type), previously reported by Noreus et al. [85Nor]. The MgNi6Si6 phase, which was also previously reported [81Buc] was not observed at the corresponding composition in the present work. However, the MgNi6Si6 phase reported as being of hexagonal symmetry (Cu1Tb type) with the lattice parameters: a= 0.4948nm and c = 0.3738nm. possibly corresponds to the μ phase (Mg(Sio.4sNio.s2h) discovered in the present work. The lattice structure of the newly discovered ro phase ((Mg0.s2Nio.4s)7S4) was determined with the help of the X-ray indexing program TREOR to be a hexagonal structure of the Ag7Te4-type with the lattice parameters, a= 1.3511nm and c = 0.8267nm.

The fracture behaviour and fracture toughness of binary and ternary intermetallic phases and composites using chevron-notched bend specimens (CNB) have been studied. Single or near single phase intermetallic alloys such as 11~ NhSi, and MgNh showed low average fracture toughness values such as ~2.0 MPa.m112, ~3.0 :MPa.m112, and ~6.0 MPa.m112, respectively. However, near Ni3Si single phase alloy tested in air showed the average fracture toughness ~31.0 MPa.m112. The composite rule-of-mixture-like relationship between fracture toughness and volume fraction of toughening Ni3Si, NiJSi+(Ni(Si)) and Ni(Si) phases showed that the fracture toughness values with increasing the volume fraction of the toughening phases seem to follow similar to the lower bound of the composite rule of mixtures.

Environmental effects on fracture toughness have been investigated for single or near-single phase alloys and selected intermetallic composites. No environmental effects were observed for near-single phase Tl, single phase NhSi, and most of the selected in-situ composites. Fracture toughness of a single phase NhSi also does not seem to be affected by the test environment. However, fracture toughness of a near-single phase Ni3Si containing fine (NiJSi+Ni(Si)) mixture seems to be susceptible to test environment. This seems to be the effect of the susceptibility of the interfaces Ni)Si/Ni(Si) in the mixture to moisture-generated hydrogen.

Indentation microcracking pattern and indentation fracture toughness of binary and ternary intermetallic phases in the Ni-Si-Mg system were studied. It is shown that the determination of the crack system as being either Palmqvist or halfpenny by simple polishing away of the indented surface is unreliable due to the existence of the core zone (crack free zone) with compressive stresses. In general, the existence of the indentation core zone in the pseudo halfpenny cracks does not seem to change the crack length-load characteristic of the halfpenny cracks allowing the use of existing equations for the penny shaped crack system to calculate indentation fracture toughness. However, equally reasonable indentation fracture toughness values are also obtained by using Shetty et al. [85She1], based on the Palmqvist crack system, which is modified in the present work. Our modification takes into account the indentation size effect (ISE) and yields results of Krc independent of indentation loads. Comparing the fracture toughness values obtained by indentation method (l.3-1.8:rvfila.m112) with those obtained by bulk CNB specimens (1.7MPa.m112) for the 11 phase, the indentation fracture toughness values are in a good agreement with those obtained on the bulk materials.