Changlin Liu

Ultra-precision machining (UPM) is an advanced manufacturing technology used to produce components with exceptionally high dimensional accuracy, surface integrity, and form precision

Brittle materials such as single crystals, polycrystalline ceramics, and amorphous glass are indispensable in modern industry. Driven by improvements in equipment performance, the required fabrication precision for optical elements and devices has reached nanoscale and is steadily advancing toward atomic level. Despite their outstanding physical and chemical properties, fabricating a defect-free surface with nanometer-level roughness on brittle materials is challenging due to microcracking, brittle fracture and severe tool wear. In recent years, field-assisted cutting has emerged to overcome the bottleneck in ultra-precision cutting of brittle materials. This review summarizes investigations of material removal mechanisms of brittle materials in ultra-precision cutting and surveys representative field-assisted cutting technologies—including laser, vibration, magnetic field, and ion implantation assisted cutting—highlighting how these fields broaden ductile-regime machining and suppress the machining-induced defects. This review further discusses the emerging multi-field coupling strategies and outlines future research directions in machining mechanisms to enable high-efficiency, low-damage, and high-consistency manufacturing of brittle materials.

Silicon wafers are the most common semiconductor substrate, and their quality plays a decisive role in the service performance of semiconductor devices. Diamond wire sawing is the first and key machining process for manufacturing silicon wafers. However, the process faces a challenge in achieving high efficiency and high quality due to the brittle characteristics of silicon wafers, which unavoidably generate surface defects like scratches and pits during the wire sawing process. In this paper, a novel self-supervised Siamese transformer (S³Transformer) model is proposed to segment silicon wafer surface defects which includes two modules, namely Siamese representation learning (SRL) and multi-scale representation learning (MRL). MRL is responsible for learning different scales of features and SRL serves as an auxiliary network to improve the small-scale feature learning by adding a Siamese loss. In addition, a novel CNN-based projector in the SRL module is designed for computation cost saving and deeper feature learning. The surface segmentation dataset of silicon wafers is established to validate the model. The results show that S³Transformer yields better silicon wafer surface defect segmentation performance compared to other state-of-the-art models.

The grain size effect is an important factor in determining the material removal behavior of polycrystalline silicon (p-Si). In the present study, to improve the understanding of nanoscale machining of p-Si, we performed molecular dynamics simulation of nanometric cutting on a p-Si workpiece and discussed the grain size effect on material removal behavior and subsurface damage formation. The simulation results indicate that when cutting on the polycrystal workpiece, the material removal process becomes unstable compared with single crystals. Higher removal efficiency, less elastic recovery and higher frictional coefficient are observed as the average grain size decreases. In the subsurface workpiece, when the grain size decreases, slip along grain boundaries merges as a nonnegligible process of the plastic deformation and suppresses the elastic deformation ahead of the cutting tool. It is also revealed that when cutting on a polycrystal workpiece with smaller grains, the average stress decreases while the workpiece temperature increases due to the impediment of heat transfer by grain boundaries. These results could provide a fundamental understanding in the material deformation mechanism of p-Si during nanoscale machining.

Cadmium telluride (CdTe) is known as an important semiconductor material with favorable physical properties. However, as a soft-brittle material, the fabrication of high-quality surfaces on CdTe is quite challenging. To improve the fundamental understanding of the nanoscale deformation mechanisms of CdTe, in this paper, MD simulation was performed to explore the nano-grinding process of CdTe with consideration of the effects of grain size and grinding depth. The simulation results indicate that during nano-grinding, the dominant grinding mechanism could switch from elastic deformation to ploughing, and then cutting as the grinding depth increases. It was observed that the critical relative grain sharpness (RGS) for the transition from ploughing to cutting is greatly influenced by the grain size. Furthermore, as the grinding depth increases, the dominant subsurface damage mechanism could switch from surface friction into slip motion along the directions. Meanwhile, as the grain size increases, less friction-induced damage is generated in the subsurface workpiece, and more dislocations are formed near the machined groove. Moreover, regardless of the grain size, it was observed that the generation of dislocation is more apparent as the dominant grinding mechanism becomes ploughing and cutting.