INTERNATIONAL JOURNAL OF MECHANICAL SCIENCES, cilt.43, ss.2699-2713, 2001 (SCI İndekslerine Giren Dergi)
Finite element studies of machining are becoming ever more sophisticated. A basic approach which removes the need, in an elastic-plastic analysis, to follow the development of chip formation from initial contact between work and tool, is the iterative convergence method (ICM). It develops a steady-state chip formation from an initial state of a fully formed chip loaded against a tool. It relies for its accuracy on the assumption that its simplified loading path coincides with the real developed flow at the end of the simulation. This paper examines the robustness of this assumption by studying the sensitivity of the simulation to changes of detail, within the ICM method, of how the flow develops; and it compares the simulated results with experiments. The experiment involves the turning of three free cutting steels, for which experimental flow stress variations with strain, strain rate and temperature, as well as information about the friction interaction between chip and tool, are available. The changes to the simulation method considered here are the structure of the finite element mesh, the measures of judging when the flow is fully developed, how the chip separates from the work at the cutting edge and the friction laws used during the approach to fully developed flow. It is shown that these do affect the outcomes of the simulation but within the ranges studied only to a minor extent and good agreement with experiment is achieved. (C) 2001 Elsevier Science Ltd. All rights reserved.