• PDF / 3,122,649 Bytes
• 6 Pages / 612 x 792 pts (letter) Page_size
• 71 Downloads / 181 Views

DOWNLOAD

REPORT

---

I. INTRODUCTION

IT has been widely recognized that the presence of a surface oxide makes aluminum difficult to sinter. The thickness of the oxide is dependent on the temperature at which it is formed and the atmosphere in which it is stored, particularly the humidity. Fresh oxide on bulk aluminum at room ˚ thick.[1–4] The temperature is reported as being 10- to 20-A thickness on atomized powder can vary from 50 to 150 ˚ .[5–8] The oxide prevents solid-state sintering in low meltA ing point metals,[9] including aluminum.[10] This has been explained in terms of the relative diffusion rates through the oxide and the metal, for metals with stable oxides.[11,12,13] Magnesia has a lower free energy of formation than alumina[14] and magnesium metal can partially reduce the Al2O3 to form spinel, MgAl2O4[15–21] by the reaction 4Al2O3 ⫹ 3Mg → 3MgAl2O4 ⫹ 2Al

[1]

This ruptures the oxide which exposes the underlying metal and facilitates sintering.[22,23,24] As little as 0.15 pct Mg is all that is required. Magnesium is therefore used to offset the detrimental influence of the oxide layer. Because of the thermal stability of the oxide, nonoxidizing atmospheres are considered to be inert during the sintering of aluminum. However, there is some literature[25] and much anecdotal evidence that indicates that sintering in nitrogen is more effective than in other atmospheres. This suggests that the atmosphere plays an active role in the sintering of aluminum. Here, we examine the sintering of aluminum in argon and nitrogen and suggest why nitrogen is a more effective sintering atmosphere.

II. EXPERIMENTAL Three systems were examined: Al, Al-0.1Mg, and Al12Sn (all compositions in wt pct). The details of the powders

G.B. SCHAFFER, Reader, and B.J. HALL, Postgraduate Student, are with the Division of Materials, School of Engineering, The University of Queensland, Qld 4072, Australia. Contact e-mail: g.schaffer@ minmet.uq.edu.au Manuscript submitted October 29, 2001. METALLURGICAL AND MATERIALS TRANSACTIONS A

are given in Table I. Powders were mixed for 20 minutes in a Turbula mixer (Glenn Mills Inc., Clifton, N.J.), and then pressed in a floating die at pressures from 50 to 400 MPa. The die wall was lubricated with a solution of stearic acid in acetone; no ad-mixed lubricant was used. Samples were sintered in a tube furnace under flowing nitrogen or argon, which had been passed over warm steel wool and through a desiccant of silica gel and magnesium perchloride to produce a dew point better than ⫺40 ⬚C. The flow rate was 1.4 ⫾ 0.1 L min⫺1. Samples of cross section 9 mm ⫻ 3.5 mm were placed on their side in a stainless steel boat, with a minimum of 2 mm separating adjacent samples to allow gas flow between them. When pure aluminum was sintered, a fresh furnace tube was used to prevent contamination, particularly by magnesium. When required, magnesium filings were placed 5 mm upstream of the sample. Samples were heated at 5 ⬚C min⫺1 to 620 ⬚C, held isothermally for 60 minutes, and air-cooled. The density of oil impregnated specimens