In this work, a predictive machining theory, based on a finite element model, were applied to dry orthogonal cutting of aluminum alloy AA2024–T351. In order to analyze the effects of chip formation process and local friction coefficient on the thermomechanical load along the tool rake face and the round cutting edge, an Arbitrary Lagrangian Eulerian (ALE) model was developed. To validate the present model, the FE results have been compared to the experimental data for a wide range of cutting speed. This shows a good agreement in both trends and values for cutting forces and tool–chip contact length. The ALE approach appears as an appropriate formulation to reproduce: (i) the tribological conditions corresponding to large values of friction coefficient such as in dry machining of aluminum alloys, and (ii) the material flow process around the round cutting edge. It was observed that a transition from a sliding contact to a sticking–sliding contact occurs when the local friction coefficient and the thermal softening are large enough. Predicted results show also that the decrease of the cutting forces as the cutting speed increases is mainly due to the variation of the tool–chip contact length in terms of cutting velocity. The effect of material flow around the round cutting edge on the distributions of frictional stress and pressure has also been analyzed.