Analysis of multicomponent evaporation in electron beam melting and refining of titanium alloys
- PDF / 1,032,771 Bytes
- 13 Pages / 612 x 792 pts (letter) Page_size
- 41 Downloads / 195 Views
I.
INTRODUCTION
ELECTRON-beam cold-hearth melting and refining (shown schematically in Figure 1) is a scrap consolidation and remelting process that has several important properties which promote production of very clean metal. Exposure to vacuum (typically 0.1 to 10 Pa) results in vaporization of volatile impurities, and the melt-flow conditions in the hearth lead to a long residence time, which promotes flotation and dissolution of inclusions. In addition, the ability to precisely control heat flux through the melt surface gives this process the potential for control of melt-flow patterns and structure in a continuously solidifying ingot.[1] Although electron beams enjoy widespread use in the production of commercially pure (CP) titanium,[2,3] their use in producing titanium alloys has been limited by poor control of composition, which necessitates subsequent ingot homogenization, usually by vacuum arc remelting.[4] This is due in part to irregularities in feed chemistry, which can be adjusted on-line, but also to frequent freezing and remelting of metal in the hearth and changes in throughput rate, which are very difficult to control. If a mechanism can be found to compensate for these phenomena, it may be possible to control composition within the electron-beam process and eliminate the need for further homogenization. Manipulation of the beam-scan frequency, whose effect on evaporation rate has been extensively demonstrated experimentally,[5,6,7] may prove to be a suitable control mechanism. As scan frequency decreases, longer beam dwell time leads to higher local superheat and increased evaporation of elements with high vapor pressure, such as aluminum in titanium. A model of this process would therefore be beneficial both to evaluate the potential use of scan frequency in composition control and, if it appears promising, A. POWELL, Graduate Student, and U. PAL, Associate Professor, are with the Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139-4307. J. VAN DEN AVYLE, Senior Technical Staff, Materials Processing Department, and B. DAMKROGER, Manager, Materials Joining Department, are with the Sandia National Laboratories, Albuquerque, NM 87185. J. SZEKELY, formerly Professor, Department of Materials Science and Engineering, Massachusetts Institute of Technology, is deceased. Manuscript submitted September 10, 1996. METALLURGICAL AND MATERIALS TRANSACTIONS B
to aid in the development of a suitable on-line control system. To date, most models of this process have assumed steady heat input from the beam.[8–11] Only the model of Nakamura and Mitchell has dealt with the effect of scan frequency on evaporation,[12] but it was applicable only to circular beam patterns at very low frequencies. A necessary parameter for the estimation of differential evaporation losses is the activity of species in the melt. The Al-Ti system has been characterized by Desai[13] and Kattner et al.[14] based in part on melt data from the work of Esin et al.[15,16] The enthalpy
Data Loading...
