A COMPARISON OF SMOOTHED PARTICLE HYDRODYNAMICS (SPH) AND COUPLED SPH-FEM METHODS FOR MODELING MACHINING

Smoothed Particle Hydrodynamics (SPH), a particle-based, meshless method originally developed for modeling astrophysical problems, is being increasingly used for modeling fluid mechanics and solid mechanics problems. Due to its advantages over grid-based methods in the handling of large deformations and crack formation, the method is increasingly being applied to model material removal processes. However, SPH method is computationally expensive. One way to reduce the computational time is to partition the domain into two parts where, the SPH method is used in one segment undergoing large deformations and material separation and in the second segment, the conventional finite element (FE) mesh is used. In this work, the accuracy of this SPH-FEM approach is investigated in the context of orthogonal cutting. The high deformation zone (where chips form and curl) is meshed with the SPH method, while the rest of the workpiece is modeled using the FE method. At the interface, SPH particles are coupled with FE mesh for smooth transfer of stress and displacement. The boundary conditions are applied to tool and FE zone of the workpiece. For comparison purposes, a fully-SPH model (workpiece fully discretized by SPH) is also developed. This is followed by a comparison of the results from the coupled SPH-FE model with the SPH model. A comparison of the chip profile, the cutting force, the von Mises stress and the damage parameter show that the coupled SPH-FE model reproduces the SPH model results accurately. However, the SPH-FE model takes almost 40% less time to run, a significant gain over the SPH model. Similar reduction in computation time is observed for in a micro-cutting application (depth of cut of 300 nm). Based on these results, it is concluded that coupling SPH with FEM in machining models decreases simulation time significantly while still producing accurate results. This observation suggests that three-dimensional machining problems can be modeled using the combined SPH-FEM approach without sacrificing accuracies.