Numéro
J. Phys. IV France
Volume 120, December 2004
Page(s) 421 - 428
DOI http://dx.doi.org/10.1051/jp4:2004120048


J. Phys. IV France 120 (2004) 421-428

DOI: 10.1051/jp4:2004120048

Computational modeling of electroslag remelting processes

K.M Kelkar1, J. Mok1, S.V. Patankar1 and A. Mitchell2

1  Innovative Research Inc, 2520 Broadway St NE , Suite 200, Minneapolis, MN 55413, USA
2  Advanced Materials Laboratories, University of British Columbia, Vancouver BC V6T 1Z4, Canada

kelkar@inres.com
alecm@interchange.ubc.ca

Abstract
Alloys used for the production of rotating components in aeroengines and land-based turbines are subject to stringent requirements to ensure absence of melt-related defects such as inclusions and segregation. Accordingly, the production of the superalloys alloys used in these applications involves multiple remelting stages, each of which plays a distinct role in ensuring that the final ingot is defect-free. Because of the complexity of these processes, high-temperature environments, and high initial and operating costs, trial-and-error based approaches for process design are inadequate. Computational modeling provides fundamental understanding of the physical phenomena and quantitative information about the effects of process parameters. Therefore, such models are very useful for design of new processes and optimization of existing processes. The paper describes a generalized framework for the modeling of the Electro-Slag Remelting (ESR) process. The model accounts for electromagnetic, fluid flow and heat transfer phenomena in a coupled manner for axisymmetric, steady-state conditions. A control-volume based computational method is used for the solution of the governing equations. The model incorporates a number of physically motivated computational features for efficient and accurate analysis of the transport processes. These include use of the effective viscosity approach for handling the liquid, mushy, and solid regions, implicit treatment of the interaction at the slag-metal interface, and contact heat transfer at the ingot-mold interface. Further, the computational method has been enhanced to address the AC electromagnetics in the ESR process. Thus, the model is able to predict the Joule heating within the slag, the distribution of the Lorentz force, the pool shape, and the motion in the slag and metal pools that arises due to buoyancy and Lorentz forces. The model is being validated using available experimental measurements for pool shape in full- scale ESR furnaces. Results of the model predictions for the flow, temperature, and electromagnetic fields are presented along with a comparison of the predicted and measured pool shapes.



© EDP Sciences 2004