Critical cooling rate and thermal stability of Zr–Ti–Cu–Ni–Be alloys

Abstract

The critical cooling rate as well as the thermal stability are measured for a series of alloys in the Zr-Ti-Cu-Ni-Be system. Upon cooling from the molten state with different rates, alloys with compositions ranging along a tie line from (Zr 70 Ti 30 ) 55 (Ni 39 Cu 61 ) 25 Be 20 to (Zr 85 Ti 15 ) 55 (Ni 57 Cu 43 ) 22.5 Be 27.5 show a continuous increase in the critical cooling rate to suppress crystallization. In contrast, thermal analysis of the same alloys shows that the undercooled liquid region, the temperature difference between the glass transition temperature and the crystallization temperature, is largest for some compositions midway between the two endpoints, revealing that glass forming ability does not correlate with thermal stability. The relationship between the composition-dependent glass forming ability and thermal stability is discussed with reference to a chemical decomposition process. 3 GFA ͑represented by critical cooling rate͒ scales with the reduced glass transition temperature T rg defined as the glass transition temperature T g divided by the liquidus temperature T l . This correlation has been confirmed in many experiments ͑see Ref. 4 for summary͒. Thermal stability in metallic glasses is usually quantified by measuring the temperature difference ⌬T between the glass transition and the first crystallization event upon heating at a constant rate. For some systems, it has been demonstrated that larger values of ⌬T tend to be associated with lower values of critical cooling rate R c . 5,6 As a result, the thermal stability has served as an indicator of GFA in these alloys.In recent years, the crystallization of Vit1 has been extensively examined.7-11 Several studies of this alloy have revealed a tendency to undergo chemical decomposition in the undercooled liquid, 7-10 which has a direct influence on the subsequent nucleation and growth of crystalline phases. Since the decomposition occurs at a temperature close to T g , the isothermal crystallization behavior of Vit1 for low undercooling is quite different from its behavior when deeply undercooled.11 In addition, a study by Schroers et al. 12 involving constant heating and cooling experiments has shown that Vit1 crystallizes in a different manner upon heating from the amorphous solid than upon cooling from the melt. In this study, a cooling rate of approximately 1 K/s was required to bypass crystallization during cooling, whereas a heating rate of 200 K/s was necessary to avoid crystallization of a detectable volume fraction. Chemical decomposition has also been observed in a series of alloys which lie along the tie line between ͓Vit1͑-a͔͒, Vit1, Vit1a, Vit1b, Vit1c, and Vit4, i.e., alloys with xϭ13-19, respectively. In order to determine the GFA of the alloys, continuous cooling rate experiments have been performed to establish R c for each. These results are compared with ⌬T and T rg values obtained through calorimetric methods, and a chemical decomposition process in the undercooled liquid is used to explain how a disparity betw...

Gold based bulk metallic glass

“…In both cases, a strong dependence of ⌬T and d c on the composition variations can be observed. However, no obvious correlation between ⌬T and d c , as previously reported to be positive 22,23 or negative, 20,24 was observed.…”

Timescales of crystallization and viscous flow of the bulk glass-forming Zr-Ti-Ni-Cu-Be alloys

“…18 Though peaks at positions corresponding to the quasicrystalline phase are still present as well, it cannot be unambiguously determined whether the quasicrystalline phase remains after the second crystallization event. There is a difference between the alloys in the types of 15 Because each sample is already partially crystallized before the constant heating rate measurement used to measure T g , the endothermic heat-flow signal during the glass transition is smaller in magnitude and it is more difficult to determine the onset, resulting in the substantial scatter shown.…”

“…The hightemperature crystallization behavior ͑above the isothermal annealing temperature at which crystallization is most rapid͒ for this alloy series has also been explored previously. 15 In this case, crystallization upon cooling at a constant rate resulted in the determination of the critical cooling rate R c , and measurements revealed that R c is lowest for Vit1͑-b͒-Vit1, i.e., these alloys are the best glass formers in the series. FIG.…”

“…A related result was observed in previous measurements performed on a differential thermal analyzer ͑DTA͒ with a constant heating rate: ⌬T was smallest for Vit1͑-b͒-Vit1 and reached its maximum value in Vit1c. 15 For all the alloys, crystallization occurs in several steps. Following the primary crystallization event, a substantial slowdown in kinetics is observed, and the second crystallization event occurs at a much later time.…”

“…However, among the more complex multicomponent bulk metallic glasses have been found several ''rogue'' glasses with glassforming abilities that do not fit the predictions of the Turnbull criterion. 10 15 These alloys are distinguished from each other by their critical cooling rate R c and thermal stability, represented by ⌬T, the temperature interval between the glass transition and the onset of crystallization. In the past, both of these parameters have been commonly and interchangeably used to describe the relative GFA for a large number of metallic glass compositions from multiple alloy families.…”