CNN is one of the deep learning technologies useful for image-based pattern recognition and classification. For machining processes, this technique can be used to predict machining parameters and surface roughness. In electrical discharge machining (EDM), the machined surface is covered with many craters, the shape of which depends on the workpiece material and pulse parameters. In this study, CNN was applied to predict EDM parameters including capacitor, workpiece material, and surface roughness. After machining three metals (brass, stainless steel, and cemented carbide) with different discharge energies, images of machined surfaces were collected using a scanning electron microscope (SEM) and a digital microscope. Surface roughness of each surface was then measured. The CNN model was used to predict machining parameters and surface roughness.

There are various micromachining processes available for manufacturing highly integrated and precise parts, each having its own characteristics and limitations. The degree to which micromachining processes meet the requirements depends on characteristics of parts that are different, making it difficult to determine the most appropriate process. In this context, the present study presents an algorithm for determining the optimal micromachining process by applying the Fuzzy AHP-TOPSIS technique frequently used for multi-criteria decision-making. Fuzzy AHP was employed for calculating weights of requirements for a given part. Fuzzy TOPSIS was employed for determining ranks of candidate processes based on weights of requirements and evaluation of processes. Fuzzy logic was applied to handle ambiguous and inaccurate information encountered in evaluating the relative importance of requirements and performances of processes. The case study in which the optimal process for micro-hole drilling of a fuel injection nozzle was determined showed that the proposed method was effective. It could be extended to micromachining of various shapes.

Microchannels machining can be used to make micro molds for microfluidic chips. The fluid flow in the channel can be controlled, by changing the cross sectional shape of the channel. V-shaped channels with a specific angle are not easily made with the etching process. This study presents the mechanical machining of microchannels of V-shaped cross section, on cemented carbide (WC-Co). In this study, to reduce tool wear in the process of machining, the micro conical tool was fabricated using polycrystalline diamond (PCD). The tool wear of the conical tool and form accuracy of channels, were investigated during V-shaped microchannel machining.

Micro needle arrays have been widely used within medical contexts, such as for drug delivery and nervous electrodes, and its fabrication methods have attracted attention recently. This paper describes a novel fabrication method for micro needle arrays on a tungsten metal plate using only laser beam machining. Tungsten material was ablated as combining laser beam scanning path to control needle structure. Overlapping laser beams from the combined scanning path causes thermal accumulation phenomena which removes the material. In this paper, controlling laser scanning path (Line Interval, Line Space, Number of Lines) and machining conditions (Power, Number of Scan), and the changing length and pitch of the micro needles were measured through SEM. Moreover, the piercing force for application in drug delivery and nervous electrodes was measured with a load cell system.

A femtosecond laser is used in various fields such as microscale machining, OLED repair, micro 3D structure fabrication, and eye surgery. Particularly, because of non-thermal property, ablation and ablation threshold are the most representative characteristics of femtosecond laser. The ablation system is accompanied by many optics, stage, or gantry. In the case of the gantry, an ordinary optic system delivers a beam where mirrors and lens are required. If the gantry moves to the sample, external stimulation such as vibration will occur. Vibration has an influence on optics such as transforming beam path and becomes an error that decreases accuracy, precision, and spatial resolution. Generally, Fiber Optic Beam Delivery System (FOBDS) is used to solve this issue. But in the case of the femtosecond laser, FOBDS is incompatible. Recently, another FOBDS model that is compatible with the femtosecond laser has been developed. In this paper, the ablation threshold was obtained by FOBDS and femtosecond laser. The results present a possibility of ablation without external stimulation.