|
|
|
Modeling Hot Tearing in the Direct Chill Casting Process
02-06-03
The direct chill, or DC, casting process was developed in the 1930's, and is considered to be one of the "top ten" most significant technical innovations for aluminum. Even with the progress in continuous casting processes such as roll or slab casting, DC casting remains a mainstay for producing the starting stock for subsequent rolling, extrusion, and forging operations that result in products for the aerospace, automotive, and packaging markets in particular. Improving the productivity of the process through a better understanding of the mechanisms involved is the focus of efforts worldwide.
The basic process of semi-continuous DC casting appears to be a straightforward one: produce a shell of solid metal in the casting mold that will contain the remaining unsolidified alloy upon exiting the mold, enabling subsequent cooling by water sprays. In reality, it is a complex interaction of a number of process variables that is further complicated by a variety of ingot shapes and aspect ratios, alloys, and product requirements. From a productivity standpoint, the casthouse would like to produce the largest possible ingots at the highest possible casting speeds. Again reality intervenes, with ingot size and aspect ratio (ingot thickness/ingot width) as well as casting speed restricted by the occurrence of casting defects, such as hot tears. Hot tearing is defined as the formation of a macroscopic separation in a casting as a result of distortion due to differential contraction of the ingot during solidification. Hot tears must be controlled to ensure that they do not propagate and result in the total failure of the ingot shell, known as a bleed out, which has significant safety ramifications.
From a technical standpoint, there is a general understanding that hot tearing is a brittle interdendritic fracture, resulting from the presence of low melting point solute rich liquid present at the grain boundaries and driven by the buildup of stresses caused by the coherent metal grains. Qualitatively it is known that alloys with larger solidification ranges, higher eutectic liquid contents, and larger cast grain sizes are more prone to hot tearing. The specific quantitative details, and especially analytical models that incorporate these details and relate them to production ingot casting operations, are just now emerging.
One extensive program in this area was the recently completed EMPACT project, a four-year BRITE/EURAM 3 effort lead by Koninklijke Hoogovens NV and including other major European aluminum producers, leading institutes, and universities. The project description states "research is proposed in order to obtain a method to control and minimize formation, cracking, and micro/macro-segregation developments in aluminum casting of slabs and billets. This is to be accompanied by identifying basic mechanisms and quantifiable process parameters, implementation of those in numerical models, and developing predictive modeling tools. An essential step is the combination of the actual experimental and production data with computational results in order to develop predictive tools for DC casting control." A number of research papers have been published from this work, including the development of a model coined the "RDG hot tearing criterion" (Drezet and Rappaz, Light Metals 2001, published by TMS, 2001, p. 888) that has been implemented in a finite element model and successfully used to predict hot tearing tendency in an Al-4.5% Cu binary alloy.
A novel testing method for assessing hot tearing tendency of aluminum alloys has been developed by McGill University and Alcan that allows tensile testing of the solidifying metal shell (Langlais and Gruzelski, Materials Science Forum, vols. 331-337 (2000), pp. 167-172). This method produces a hot tearing susceptibility value, which is the inverse of the maximum tensile stress that can be supported, and allows ranking of commercial aluminum alloys in terms of hot tearing susceptibility that is consistent with that found in practice. This technique has been coupled with a DC casting surface simulator to evaluate the role of mold surface roughness and its influence on cast grain structure and subsequent hot tearing tendency in an Al-4.5% Cu alloy (Fortier, et al., Light Metals 2001, published by TMS, 2001, pp. 839-846).
The ability to better understand hot tearing and develop methods to control it
could result not only in improved productivity but also reduced scrap and attendant
savings in energy. It is this latter aspect that has made this area one of interest
to the Department of Energy's Office of Industrial Technologies Aluminum Industries
of the Future program. They are supporting a program led by Secat, Inc. entitled
"Modeling and Optimization of Direct Chill Casting to Reduce Ingot Cracking".
Involving researchers from three major National Laboratories as well as aluminum
producers and an ingot casting equipment supplier, the goal of this project is
"to assist the aluminum industry in reducing the incidences of stress cracks from
5% down to 2%." The research focus is on "developing a detailed model of heat
conditions, microstructure evolution, solidification, stress/strain development,
and crack formation during the DC casting of aluminum. This model will provide
insights into the mechanisms of crack formation, butt deformation, and aid in
optimizing DC process parameters and ingot geometry." The project is roughly halfway
through its three-year span, and so far emphasis has been on collection of temperature
and heat transfer data on commercial scale ingots and initial model formulation.
Further information is available on this project at http://www.eere.energy.gov/industry/aluminum
Two recent DC casting developments highlight the importance of understanding the process analytically. The successful use of a combined modeling and experimental approach by Corus Aluminum Rolled Products to produce 5083 ingot of 600 x 3000 mm in cross section was reported recently (Grealy, et al., Light Metals 2001, published by TMS, 2001, pp. 805-811 and pp. 813-821). Another example is the development of shaped ingot cross section by the DC casting process, called NetCastÔ, by Wagstaff, Inc. for use as preform stock for forging. Using an understanding of the conditions for crack formation and adjusting cooling accordingly, a shaped product with a fine cast grain size, resulting in forged properties superior to those from a conventionally cast and extruded stock, was reported (Anderson, et al., Light Metals 2001, published by TMS, 2001, pp. 847-853). These applications show how valuable a good fundamental understanding of the DC casting can be, and bode well for further improvements with the new knowledge being developed in the research activities recently completed and underway.
Article provided courtesy of The Aluminum Association - www.aluminum.org
|
|
|
