Since only uni-direction motion is produced by traditional abrasive flow machining (AFM), so the polishing effects of the inner hole is not easy to achieve uniform roughness of the whole surface after polishing. Therefore, in this study, a rotating device with a DC servo motor was set up in the AFM to increase the tangential forces on the machining surface, and therefore, improve the uniform surface roughness and polishing efficiency. The rotating device was designed by a group of transmission gear set and a DC servo motor to create a rotational finishing path for the abrasive medium. The rotational motion of an abrasive can create different tangential forces on the working surface, inducing a more complex polishing path than that of traditional AFM. In addition to rotational speed, a servo motor can also change rotation directions in one working process, causing an abrasive medium to create many irregular finishing paths in the AFM. The experimental results showed that the surface roughness of the workpiece was significantly decreased with an increase in the rotational speed. Additionally, the results also showed that the surface roughness (SR) of the inner hole decreased from 0.61 μm Ra to 0.082 μm Ra after 20 machining cycles, the surface roughness improvement rate reached 87% at 15 rpm rotational speed, by applying a 1.5:1 silicone gel/abrasive concentration ratio and #60 abrasive mesh in the experiments. This study created excellent polishing efficiency by using a servo rotational device with AFM to produce good surface quality.

This paper presents the validation of the design of a 6-in. $f/2.2$ dual-wavelength transmission sphere (TS) based on the Fizeau interferometer. The TS was verified at a wavelength of 632.8 nm and is compatible for measuring the transmitted wavefront error of an i-line lithography lens. The achromatic design is imperative for a dual wavelength TS and requires more lenses for correcting the chromatic focal shift. Because the overall weight of the TS should be controlled within the load range of the piezoelectric transducers of the interferometer, the fabrication and mounting of lenses with high aspect ratios are challengeable. The mounting of the reference surface based on three-point mounting was successful for the residual reference wavefront error under peak-to-valley (PV) $λ/10$. Furthermore, the reference wavefront is typically restricted within PV $5λ$ to avoid distorted interference fringes. Therefore, we built a double-pass interferometer model for tolerance analysis, and the error budget facilitated decision-making regarding the suitable specifications of lens manufacturing and assembly for cost efficiency. The test results demonstrated that the deformation of the reference wavefront and the residual reference wavefront error met the critical specification at 632.8 nm and that the achromatic TS is compatible for measuring i-line lithography lenses.

In the era of Industry 4.0, numerous AI technologies have been widely applied. However, implementation of the AI technology requires observation, analysis, and pre-processing of the obtained data, which takes up 60–90% of total time after data collection. Next, sensors and features are selected. Finally, the AI algorithms are used for clustering or classification. Despite the completion of data pre-processing, the subsequent feature selection and hyperparameter tuning in the AI model affect the sensitivity, accuracy, and robustness of the system. In this study, two novel approaches of sensor and feature selecting system, and hyperparameter tuning mechanisms are proposed. In the sensor and feature selecting system, the Shapley Additive ExPlanations model is used to calculate the contribution of individual features or sensors and to make the black-box AI model transparent, whereas, in the hyperparameter tuning mechanism, Hyperopt is used for tuning to improve model performance. Implementation of these two new systems is expected to reduce the problems in the processes of selection of the most sensitive features in the pre-processing stage, and tuning of hyperparameters, which are the most frequently occurring problems. Meanwhile, these methods are also applicable to the field of tool wear monitoring systems in intelligent manufacturing.