Search

Metal cutting is nowadays mostly modelled with the Lagrangian formalism when the finite element method is adopted. With these models, the results depend on the mesh parameters, such as the elements size, orientation, etc. While this is a known problem, few studies that analyse it systematically and quantitatively can be found. The Coupled Eulerian Lagrangian (CEL) formalism is an attractive alternative to the Lagrangian formulation. The mesh dependency issue has still to be highlighted when CEL is adopted. The aim of this paper is to study the mesh influence on the results of a CEL model and to provide recommendations for future modelling. A benchmark is carried out based on an experimental reference in strictly orthogonal cutting conditions. The comparison of the numerical results with this reference shows that the size of the elements influences significantly the chip morphology and the cutting forces. This size should be at least two times smaller than the cutting edge radius of the tool to obtain results with an acceptable accuracy. The orientation of the elements is an important parameter as well. It is such that it is recommended to use square elements to avoid any influence of the elements orientation on the results. On the contrary, the results are independent of the elements size and number in the direction perpendicular to the cutting plane. For the cutting conditions of this study, square elements of 5 μm size should be used to obtain converged numerical results. [ABSTRACT FROM AUTHOR]

A wide range of parameters is available in the literature for the Johnson-Cook constitutive model, the most used flow stress in metal cutting modelling. This strength of the Johnson-Cook model is also a weakness as it is therefore difficult to choose the set that will provide the most accurate results. Indeed, for a given material, several parameters are available due to different identification conditions (strain, strain rates and temperature range). This paper gathers the sets of Johnson-Cook parameters from the literature for Ti6Al4V before comparing them based on their stress-strain curves. The twenty identified sets are then introduced in a Coupled Eulerian-Lagrangian (CEL) finite element orthogonal cutting model and their results are compared at fixed cutting conditions. This comparison is then extended for the three most characteristic sets to two other uncut chip thicknesses. All the numerical results are compared to an experimental reference in the same cutting conditions. Links between the parameters of the Johnson-Cook flow stress and outputs of the model are then highlighted; they are still valid when the uncut chip value is modified. A set is finally recommended to obtain an accurate estimation of the cutting force and average feed force and chip thickness values. [ABSTRACT FROM AUTHOR]

Numerical modeling of metal cutting is a complex problem involving many nonlinear and coupled phenomena (high speed, large strains, large strain-rates, heat generation, etc.). Due to the large deformations, the mesh of the workpiece is often highly distorted. When it does not ends prematurely the computing, it slows it down and decreases the confidence in the results. Many formalisms are encountered in the literature to model the orthogonal cutting process, each with their pros and cons, more or less significant. This paper introduces the Coupled Eulerian-Lagrangian (CEL) formulation, in which the workpiece is described by the Eulerian formulation and the tool by the Lagrangian one. The comparison with experimental results and an Arbitrary Lagrangian-Eulerian (ALE) model with pure Lagrangian boundaries shows that the chip morphology and the cutting forces are well predicted by that new model. The absence of mesh deformation in the part does not decrease the stable time increment and therefore does not slow down the computing, which is competitive compared to the reference ALE model. This also increases confidence in the results given by the CEL model and opens new possibilities in the modeling of metal cutting. [ABSTRACT FROM AUTHOR]

The titanium alloy Ti6Al4V, although widely used in machining, is a hard-to-machine material, particularly due to the production of saw-toothed (or segmented) chips. The phenomena leading to this kind of chips are still an ongoing problem in the current literature, as well experimentally as numerically. An experimental orthogonal cutting setup on a milling machine is presented as a benchmark to validate finite element models. Its ease of implementation and near-to-zero cost should enable it to be used for any material, as long as the cutting speed is achievable by the feed rate of the machine. In addition to the study of chip formation, this setup could also allow to carry out long, although interrupted, machining tests to analyze tool wear. A numerical finite element model is then introduced to model the chip formation in the same cutting conditions. The comparison of the numerical results with the experiments shows that the level of the forces is mainly influenced by the material constitutive model, while the chip morphology is mostly impacted by the chip separation criterion. Either experimentally or numerically, the mechanism leading to the formation of a Ti6Al4V saw-toothed chip, for the cutting conditions of this study, is the conjunction of the adiabatic shear band in the primary shear zone, followed by the propagation of a crack into it. [ABSTRACT FROM AUTHOR]

Machining by chips removal operations depend on a large set of parameters, which leads to a time consuming and expensive experimental optimization of this process. Numerical finite element modelling of orthogonal cutting is its most attractive alternative at this time. Apart from the difficulties caused by the complexity of the phenomena involved, the industrial application of this method comes up against unacceptable CPU computing time. The mass scaling is a numerical technique allowing to artificially speed up these calculations. This paper presents its application to a Lagrangian orthogonal cutting finite element model of the most commonly machined titanium alloy, Ti6Al4V, which has not previously been performed. Once the adaptive mass scaling is enabled, the CPU computing time is reduced by about 70% for a typical computation. This improvement should not be performed at the expense of the quality of the results by comparison to the experimental reference (chip morphology, formation mechanism, cutting forces, teeth formation frequency, etc.), nor impact significantly the numerical computation (total mass increase of the model, for example). This study shows that, when used carefully, the adaptive mass scaling constitutes an efficient method to reduce the CPU computation time. It should therefore be considered for the development of future models. [ABSTRACT FROM AUTHOR]

Abstract: The physics of formation of a saw-toothed (or segmented) Ti6Al4V chip in machining is still an ongoing issue in the literature as well as its modelling and numerical simulation. After a review of the theories explaining the formation of such a chip, three finite element models are presented with a desire to take into account the physical phenomena. The results of the simulations are compared with the different theories reported in the literature. This confrontation allows us to keep the most relevant model which contributes to the understanding of the phenomena. It shows that combining the adiabatic shear band and the crack propagation leads to a chip morphology and formation mechanism close to the experimental one. The study also allows to estimate the teeth formation frequency of the numerical chip. [Copyright &y& Elsevier]