Orthogonal cutting experiments using a quick-stop device are performed on Al2024-T3 and OFHC copper to study the chip–workpiece interface in a scanning electron microscope. Evidence of ductile tearing ahead of the tool at cutting speeds of has been found. A numerical finite element model is then developed to study the energy consumed in material separation in micro-cutting. The ductile fracture of Al2024-T3 in a complex stress state ahead of the tool is captured using a damage model. Chip formation is simulated via the use of a sacrificial layer and sequential elemental deletion in this layer. Element deletion is enforced when the accumulated damage exceeds a predetermined value. A Johnson–Cook damage model that is load history dependent and with strain-to-fracture dependent on stress, strain rate, and temperature is used to model the damage. The finite element model is validated using the cutting forces obtained from orthogonal micro-cutting experiments. Simulations are performed over a range of uncut chip thickness values. It is found that at lower uncut chip thickness values, the percentage of energy expended in material separation is higher than at higher uncut chip thicknesses. This work highlights the importance of the energy associated with material separation in the nonlinear scaling effect of specific cutting energy in micro-cutting.