Measurement of Precipitate Nucleation Times in Molten Metals by Pulsed Surface Melting
- PDF / 1,019,869 Bytes
- 5 Pages / 417.6 x 639 pts Page_size
- 14 Downloads / 169 Views
MEASUREMENT OF PRECIPITATE NUCLEATION TIMES IN MOLTEN METALS BY PULSED SURFACE MELTING*
D.M. FOLLSTAEDT, S.T. PICRAUX, P.S. PEERCY, J.A. KNAPP, Sandia National Laboratories, Albuquerque, NM 87185
AND W.R.
WAMPLER
ABSTRACT The short melt duration resulting from pulsed laser and electron beam surface melting of ion-implanted metals has been used to measure precipitate nucleation times of compounds within the melt. We have examined the phases present in several alloy systems with TEM and used calculated thermal histories to place limits on the time required for nucleation of the following compounds: AlSb (5-25 ns), Al 3 Ni (> 750 ns), Al 3 Ni 2 (> 950 ns) and AMNi (< 1000 ns), all in molten Al, and TiC (, 100 ns) in molten Fe. The compounds observed after our rapid solidification have relatively simple, cubic structures and melt congruently, while those predicted but not observed have more complex structures and decompose peritectically.
INTRODUCTION The nucleation rate of crystalline phases is an important factor in determining the microstructure of rapidly solidified alloys. By avoiding the nucleation of equilibrium phases, non-equilibrium alloys such as metastable solid solutions, new crystalline phases and amorphous phases can be formed. To obtain quantitative information on the time needed for nucleation, a quenching technique is needed which has a well characterized temperature history and a melt duration comparable to the nucleation time. Pulsed surface melting of ion implanted metals meets these criteria in the nanosecond time regime. Below we discuss the bounds on nucleation times which we have determined from transmission electron microscopy (TEM) observations on pulsed melted Al(Sb), Al(Ni) and Fe(Ti,C) alloys. In our procedure, a host metal is ion implanted with one or two other atomic species which react either with the host or with each other to form compounds which coexist with the host liquid at its melting temperature. Alloy additions up to 30 at.% are implanted to depths < 0.ilAm. Concentration profiles are adjusted by varying the ion energy and are measured by ion backscattering with a depth resolution - 0.01 pm. A pulsed laser or electron beam [1,2] is used to melt through the alloyed layer and into the pure metal substrate (,:2 pm melt depth). The resulting steep thermal gradients in the substrate then drive the liquid-solid interface rapidly back to the surface. Diffusion of the implanted species occurs within the melt, but is usually too slow for surface segregation to occur in metals [2,3]. The shorter time of our laser pulse relative to the electron pulse (20 ns versus 70 ns) and the shallower energy deposition depth ( -20 nm for X = 0.69 pjm versus -1 pm for 10-20 keV electrons) provide shorter melt times (- 50 ns versus - 500 ns), more rapid solid phase cooling rates (~ 1010 K/s versus -108 K/s) and faster interface velocities ( -20 m/s versus - 5 m/s). Thus these two directed energy sources allow precipitate formation to be studied as a function of melt time. The resolidification param
Data Loading...
