High-aspect-ratio microdrilling demonstrated by the forth harmonic of nanosecond Nd:YVO4 laser. In this paper, we investigated the depths of the holes drilled in alumina, aluminum alloy, cemented carbide, high speed steel, silicon and stainless steel. Firstly, the depths of the holes drilled with the repetition rate of 1 kHz, 5 kHz, 10 kHz and 15 kHz, were measured. As a result, the deepest holes were drilled at the repetition rate of 10 kHz in all the materials. At the tip of the holes drilled in cemented carbide, high speed steel, silicon and stainless steel were blanched. The depths of the holes were 600 μm - 1400 μm. The range of the holes with the diameters less than 10 μm were measured, and aspect ratios were calculated. As a result aspect ratios of alumina, aluminum alloy, cemented carbide, high speed steel, silicon, and stainless steel were 104, 26, 17, 56, 64 and 55, respectively. No clear redepositon layer formed on the inner surface of silicon. Scanning electron microscope observation revealed that inner surface was not flat.
This paper investigates the machining phenomena in electrical discharge machining (EDM) of carbon fiber reinforced plastics (CFRP) in order to understand its machining process. Sinking EDM using deionized water for machining liquid and wire EDM were carried out, and machining properties, such as material removal rate and electrode wear ratio, were compared with those of EDM of metal. The waveforms of discharge current and gap voltage were also observed during the machining process, and the ratio of the short-circuiting was evaluated. Experimental results show that Sinking EDM and wire EDM of CFRP can be realized without using special machining method and equipment. It is found for the sinking EDM of CFRP that electrode wear ratio decreases with increasing the discharge current, and the ratio less than 1% can be achieved even when deionized water is used for the machining liquid. It is also found that the short-circuiting caused by the contact of carbon fibers that are fluffing on the machined surface with the tool electrode frequently occurs, and the short-circuiting contributes to material removal.
This paper deals with the development of a novel polishing technique for the glass substrate for electric devices by applying AC electric field. In order to grasp the movements of the slurry under AC electric field, we have observed slurry behaviors with an observation device we developed, and found that the slurry was smoothly led to the polishing area by AC electric field. Furthermore, effectiveness of the slurry distribution on the polishing area increased by 12%. Polishing rate also increased 22% compared to the conventional polishing when AC electric field was applied. This will contribute to the reduction of the consumption of the cerium oxide abrasives.
Whole-space tabulation method(WSTM) is a technique to relate the phase of a projected grating and the coordinates at each pixel of a camera. The relation data of all the pixels are obtained in the whole-space where a reference plane is moved. The relation data are stored in tables. Thereby, WSTM is high speed and accurate shape measurement method because the coordinates are obtained by only looking up the tables without any complex calculation. Even when the brightness distribution of the projected grating is warped from cosinusoidal wave, correct coordinates are provided. In this paper, it is confirmed that accurate shape measurement can be performed by defocusing a projected grating with a rectangular brightness distribution.
It is extremely necessary to estimate the spatial locations of sound sources under various vibrations and noise phenomena. With this background, techniques for estimating the locations of sound sources as non-contact testing are widely used for understanding various vibrations and noise problems intuitively. In this study, our goal is to estimate locations of high correlation sound sources with high accuracy by combined SSP and MUSIC methods, based on the leading edge technology of radar signal processing. Considering this objective, the handy microphone array system for evaluating the performance of the combined SSP and MUSIC methods was developed experimentally. This paper presents details of the performance of the combined SSP and MUSIC methods as well as the analysis method of the measured and experimental results of the performance tests for estimating the locations of high correlation sound sources. As a result, it is possible to find the sound source locations with high sensitivity applying the combined SSP and MUSIC methods through a series of experimental and computational examinations.
In recent semiconductor industry, production and evaluation of ever flatter, thinner and larger Si wafer are required to fulfill the demands in high integration and cost reduction. However, uncertainty in measurement of wafer profile is encountered as a severe problem because the resolution required for analysis of wafer flatness is reaching the limits of available instrument. Thus, the measured profile data always incorporate a certain degree of noise and this noise component behaves to bias the outputs of wafer profile such as SFQR (Site Frontsite least-sQuares focal plane site Range), GBIR (Global Backside Ideal Range), bow, warp and etc.. In order to precisely evaluate the wafer geometry, it is essential to remove the noise from the measured data. Described in this paper is design and development of digital filters for denoising by use of wavelet transform. Compared to the conventional low-pass filters, the developed filter not only provides better performance of decomposition in the spatial frequency domain, but also offers the new capability of filtering in amplitude components.
New-OJT system performed worker's skill education at remote place via e-learning system. However, each worker's skill level education have not suitable arrangement method based on different individual workers' knowledge. Therefore, this paper aims to propose a support method for constructing the training content of individual skill education of New-OJT system. Bayesian network was applied for creating the skill training content from the obtained NC training operation data, successful training course that was analyzed and predicted by text mining and neural network technique, respectively. The results of experiments revealed that the proposed method can achieve increasing number of workers for individual workers' skill level of New-OJT system.
This paper deals with the development of a simulator applying mixed reality for the training of manual control milling machine operation and the evaluation of the performance. This system is mixed reality : tool and workpiece are virtual objects, and milling machine is an actual object. Therefore, in operation training, the simulator provides the same manipulative feeling as that of the actual machine to the trainee. The displacement sensors measuring the table position of the machine achieve the registration error between the tool and workpiece under 0.05 mm required for cutting simulation calculation. As the result of the operation with the simulator, it is confirmed that the simulator is effective with the high presence and safe operation environment.