In this paper, a systematic evaluation of six ductile fracture models is conducted to identify the most suitable fracture criterion for metal cutting processes. Six fracture models are evaluated in this study, including constant fracture strain, Johnson-Cook, Johnson-Cook coupling criterion, Wilkins, modified Cockcroft-Latham, and Bao-Wierzbicki fracture criterion. By means of abaqus built-in commands and a user material subroutine (VUMAT), these fracture models are implemented into a finite element (FE) model of orthogonal cutting processes in abaqus/Explicit platform. The local parameters (stress, strain, fracture factor, and velocity fields) and global variables (chip morphology, cutting forces, temperature, shear angle, and machined surface integrity) are evaluated. The numerical simulation results are examined by comparing to experimental results of 2024-T3 aluminum alloy published in the open literature. Based on the results, it is found that damage evolution should be considered in cutting process FE simulation. Moreover, the B-W fracture model with consideration of rate dependency, temperature effect and damage evolution gives the best prediction of chip removal behavior of ductile metals.