Modeling of Grinding Process of Silicon Carbide Fiber Reinforced Silicon Carbide (SiCf/SiC)

This research explores the advanced modeling and optimization of the grinding process for silicon carbide fiber-reinforced silicon carbide (SiCf/SiC) composites, leveraging state-of-the-art simulation techniques. SiCf/SiC composites are valued for their exceptional mechanical properties, including high strength, hardness, and thermal resistance, making them ideal for demanding applications in the aerospace, nuclear, and automotive industries. Grinding, as the second last manufacturing process, plays an important role to achieve high dimension accuracy and ensure good surface integrity. However, their inherent brittleness and extreme hardness pose significant challenges for the manufacturing process, adding even more cost to conventional iterative testing on process parameter. Therefore, innovative and cost-effective approaches are essential to identify optimal process parameters that yield high-quality surface finishes and precise dimensional tolerances. The research begins by applying the Smoothed Particle Hydrodynamics (SPH) method to model single-grain diamond scribing of SiCf/SiC, which represents the abrasive-workpiece interaction mechanism. A generalized simulation configuration, contact criteria and material properties are validated against experimental force measurement and surface damage analyses with different scribing speed, diamond geometry, depth of cut and fiber orientations. Building on this foundation, the Smoothed Particle Galerkin (SPG) method is introduced to refine the definition of fiber-matrix interactions, and enhance the prediction of varying material removal mechanisms under different operating condition. SPG method also demonstrates its superiority in computational stability and efficiency, attributing to its smoothing algorithm, improved kernel function for severe damage, and bond-based failure model. Massively-parallel processing is then incorporated to enhance scalability of SPG model, reducing simulation times by over 90% and enabling practical application in large-scale industrial scenarios. Different domain decomposition strategies are evaluated to optimize force prediction accuracy and computational efficiency. To address the complexity of the grinding wheel topography, a statistical study on grinding wheel topography is conducted. A new method is proposed to reconstruct grinding wheel surface according to the distribution of diamonds protrusion height, location, and geometry. The multi-grit scribing model is validated through the creep-feed grinding test on SiCf/SiC, where the grinding force is successfully predicted. Material removal mechanisms, such as SiC fiber debonding, the development of lateral and radial cracks, SiC matrix fracture, and ductile deformation, are effectively captured under various process parameters. By integrating these innovations, this research provides a comprehensive model that can serve as an efficient alternative to expensive process parameter testing. These findings can advance the precise and cost-effective machining of SiCf/SiC composites, facilitating their broader adoption in critical high-performance applications.