This paper aims to design a novel pulse generator for electrical discharge machining (EDM) which can generate more efficient pulse with high frequency. The LC generator which uses an inductor to store energy and a switch to control the gap voltage and current flowing through the gap can realize higher discharge energy per pulse even with a low voltage DC power supply and obtain large removal volume per discharge due to high rising speed waveforms. However, the discharge probability of the LC generator is low because the duration of open voltage pulse is significantly short and the ignition usually occurs after a random delay time. Hence, this paper proposed a kind of the modified LC generator in which a high impedance transistor pulse generator is used to ignite discharge followed by the main discharge current supplied from the LC generator to increase the discharge frequency.

In production systems consulting, the problem-solving level required increases with the recent increase in the complexity of the client’s production systems. The quality, cost, and delivery (QCD) of consulting activities depend on consultants' skill levels. This means many unskilled consultants cannot satisfy the requirements of clients. Based on a human-dependent approach, resulting in the problem that the QCD level of consulting has not reached problem-solving. This paper proposes a method to systematize the data, information, and knowledge used by consultants by applying the concept of Digital Triplet so that they can be reused for other consulting.

In machining of rolled titanium alloy, the anisotropy in the cutting force induced by the rolling texture should be controlled to achieve high quality surface. The cutting tests on the normal direction (ND) plane of a rolled titanium alloy plate is conducted with changing the cutting direction with respect to the rolling direction. The deformation and the cutting force in micro cutting are discussed compared to those in cutting of single crystal titanium. Although the anisotropic effect on the cutting process is smaller than that of the single crystal workpiece, little deformation occurs around the boundary between the groove and the unmachined surface when cutting along the rolling direction. The material behavior is characterized by the shear plane cutting model. The shear angle and the shear stress on the shear plane depends on the cutting direction for the rolling direction.

In order to select appropriate temperature measuring points of the workpiece in end-milling process monitoring, an evaluation criterion to select measuring points had been proposed, which is based on the thermal characteristics of workpiece calculated by pre-process simulation and estimated variations. Magnitude of temperature change and sensitivity to the process variation are utilized to formulate the criterion. A systematic evaluation results of criterion under different sets for measuring points candidates were compared and characteristics of the proposed criterion were investigated. The criterion could evaluate the significance of each measuring point candidate. However, it is difficult to determine a sufficient number of measuring points.

We develop an alignment monitoring system which adopts a three-dimensional length measurement grid based on multiple absolute distance interferometers. Each interferometer refers an interval of an optical pulse train of an optical frequency comb filtered out by an etalon. The interval is determined by a free spectral range, FRS of the etalon, which should be evaluated accurately with an uncertainty better than several ppm. Here, we evaluate the FSR from a distance between interference signal peaks generated by the etalon and the optical frequency comb by referring a repetition frequency of the optical frequency comb. Then, we compensate the evaluation based on a relation between the observed peak distance change against the repetition frequency change of the optical frequency comb.

A surface texturing method using the vibration-assisted microcutting was developed. In such a process, a triangular pyramidal diamond indenter is oscillated in the cutting depth direction with an amplitude of several tens of μm using a fast-tool-servo system. The burrs or pileups generated around each dimple like pattern after texturing can be removed on-machine by just supplementary conducting the conventional microcutting. In this report, c2680 brass plate surfaces were successfully textured with a few different kind of textured area densities using the developed method. Their dry sliding friction and wear characteristics were examined using the ball-on-plate sliding tests. As a result, it was clarified that both the coefficient of friction and wear were decreased with increasing the textured area density.

The Selective Laser Melting (SLM) of the powder bed fusion is a method of laminating metal powder that are selectively melted and solidified by laser beam to build a three-dimensional object. In recent years, its use has expanded in various fields such as the automotive industry, medical industry, and aerospace industry. In this study, we propose a molten pool formation model that uses only a small number of parameters that dominate melt pool shape formation. The formula was derived and compared with the experimental values.

Microcracks are induced on a glass lens surface during machining process. It is necessary to develop a measurement method for the stress state of microcracks based on the backscattered light from microcracks. We propose an evaluation method of changes in the molecular structure caused by stresses around the cracks based on Raman spectroscopy of the backscattered light. In this report, Raman measurements were carried out at an indentation and an indentation crack. Raman spectrums changes at the sample surface of the indentation. The possibility for detection of changes in the molecular structure of glass around a microcrack from Raman measurement is confirmed.

In recent years, automation of operations in the manufacturing industry has been actively pursued to reduce production costs. As part of this eff ort, systems have been developed to automatically determine the machining sequence instead of the expert. However, it is difficult to design a rule that completely reproduces the expert's decision. In this study, we propose to acquire rules that can emulate expert decisions via inverse reinforcement learning (IRL). The developed system acquires rules by training a neural network using past manufacturing, which represents the results of the expert's decision.

In recent years, 3D nano-periodic structures have attracted much attention because the structures have a function; for example, entropic cages near a nanopore can trap DNA. To fabricate periodic structures, this paper focused on Talbot lithography which has efficiency for processing. Previously, convolutional neural network (CNN) is utilized to estimate phase and amplitude of incident light for exposurement in Talbot lithography. However, light intensity distribution estimated by the incident light is not in good agreement with design distribution because of too much information of datasets. Moreover, the estimated distribution does not have periodicity. To overcome the problems, this paper focused on CNN with binarized datasets to improve flexibility of Talbot lithography. An estimation result trained by CNN with a binarized dataset is in good agreement with design distribution. Additionally, it is possible to control local intensity distribution by controlling threshold for processing.

Effective redesign using existing products minimizes the cost of designing plastic bottles. A method that performs various deformation operations on a redesignable model, which is generated by using implicit surface modeling and frequency analysis is proposed. We converted point clouds obtained from scanning existing real products into implicit surface models and performed deformation experiments. Results show that our method easily changes polygonal shapes of the cross-section and can extract and paste patterns.

The objective of this study is to detection abrasive grains using deep learning. We have used SSD for the object detection model to detect the abrasive grains, but in this study, we chose EfficientDet, which is an advanced form of SSD. This model consists of three stages. First, a backbone network based on EfficientNet. EfficientNet is a classification model using deep learning. Second, fusion multiple scale each feature maps that named BiFPN Layer. The third is the part that classifies and locates the abrasive grains. In this paper, we changed the aspect ratio of the default box to detect abrasive grains. Finally, we succeeded in determining the most suitable aspect ratio for the detection of abrasive grains.

In electrochemical machining, the contact area with the electrolytic solution is processed, including unintended areas, and the processing accuracy deteriorates. Therefore, instead of the liquid electrolyte, electrochemical machining using quasi-solid gel impregnated with the electrolyte was proposed to improve the processing accuracy by limiting the current range. However, its processing characteristics have not yet been clarified. In this study, electrochemical machining was performed using the quasi-solid electrolyte of agar. The effects of melting of the quasi-solid electrolyte by Joule heat and kind of electrolyte on electrochemical machining were investigated and certain trends were discussed.

In order to improve the machining accuracy of electrolytic machining, an electrolytic solution suction tool with an auxiliary anode that limits the existing area of the electrolytic solution and current flow is proposed. However, there are many unclear points about the electrolyte suction tool with the auxiliary anode. In this research, we investigated the machining phenomenon by observing the flow of the electrolyte during machining and the bubbles generated during machining with a two-dimensional electrolyte suction tool with an auxiliary anode and visualized the machining state.

Electrochemical machining can be used to create fine taper holes by transferring the electrode shape to straight fine holes. However, its transfer accuracy is low, and the relationship between the electrode taper angle and the transfered hole angle is not clear. The aim of this study is to find machining parameters improving the transfer accuracy and to clarify the relationship between the electrode taper angle and the transfer taper angle. Experiments by changing the electrode taper angle, the electrolyte flow rate and the electrolyte concentration were conducted to investigate the relationship between the electrode taper angle and the transfer taper angle, and the transfer accuracy.

In order to investigate the effect of ultrafine bubbles in a coolant, the authors developed the antifoaming techniques which eliminate the gaseous bubbles from the coolant with a mixture of gaseous and oil bubbles. The effect of gaseous bubble in a coolant is evaluated by both the grinding force and the surface temperature of a workpiece during the surface grinding. Finally, the cooling efficiency of a coolant with gaseous bubbles is evaluated.

This study describes a finishing performance of cooling channels built by laser powder bed fusion (LPBF) using free abrasive grains. Different types of cross-sectional channel are built, and the effect of internal face finishing on the surface quality and cooling performance is investigated. As results, the use of free abrasive grains enabled the improvement of the cooling performance in addition to the surface quality in the cooling channel. The cooling performance was depending on the surface area of channels and the triangle cross section was suitable for the achievement of uniform surface quality due to the limitation of LPBF.

The reuse of products and materials has been attracting attention as one of the efforts to shift to a sustainable society. One of the ways to reuse products is to lease them, and the demand for leasing is increasing as a means to make effective use of resources. In the leasing industry, the number of inventories at each site depends on the experience and intuition of the person in charge, and if the number of inventories is too large, the storage cost of the inventories increases.

Therefore, this study aims to determine inventory management methods that can respond to demand fluctuations and reduce costs by using demand data and inventory movement histories obtained in the field.

This paper presents an identification method for milling simulation parameters by utilizing spindle speed variation (SSV) method. In order to realize highly-efficient and highly-accurate cutting, optimization of processing conditions is inevitable. Cutting simulation is utilized to predict cutting force and process stability. As the prediction accuracy depends on accuracy of the parameters used for the simulation, the identification is significantly important. The fast identification method developed by the authors in the previous study has a problem of low identification accuracy that should be improved. To improve the identification accuracy, SSV method is applied in the present study. Analytical investigations verified that the parameters identification accuracy is significantly improved by applying SSV.

This paper deals with job shop scheduling using priority rules. Because of its computational efficiency, applying priority rules is the most practical approach for real-world practical scheduling problems. However, effective priority rules have been developed by trial and error, and scheduling performance generated by simple priority rules is not always good. This paper examines learning priority rules for job shop scheduling with regard to weighed tardiness using the data obtained from optimal or near-optimal schedules. The optimal or near-optimal schedules are generated using an IP solver or a genetic algorithm. Numerical experiments show that it is possible to learn effective priority rules if training data of enough quality are used for learning.

Manufacturers faced severe conditions due to frequent natural disasters and spread of the COVID-19 virus need to continue the production even in any emergency situations. We proposed a reconfigurable production line composed of resources that include line workers and multipurpose equipment. Skilled workers can handle complicated tasks quickly in unexpected situation but require high production cost after transition to steady situation. Equipment takes time temporarily to launch and adjust a task but can stabilize the quality regularly. This paper describes a production line design method for maximizing production efficiency in both steady and emergency situations by modeling the capabilities of resources.

Chemical-Mechanical Polishing (CMP) process has emerged as the method of choice for planarization in the semiconductor industry. The interaction force between the surface being polished and the abrasive nanoparticles plays a significant role in flattening the substrate surface in the CMP process. For example, when the SiO₂ nanoparticle is distant from the SiC substrate surface, the interaction force between the surface and the nanoparticles is presented. This force could be analysed using Derjaguin and Landau and Verwey and Overbeek (DLVO) theory and particle Brownian motion. According to the simulation results, we found that very fine abrasive nanoparticle such as smaller than approximately 10 nm is comparatively approach to the surface being polished more certainly.

This paper presents a novel chatter simulation technique considering amplitude-dependent flank contact force, i.e., process damping force. In the conventional analysis model of process damping, the amplitude and tool shape dependencies could not be considered, and hence prediction of the resultant amplitude under the quasi-stable process is difficult in the time domain simulation. In this study, a new model utilizing finite element simulation is developed, where nonlinearity of process damping is considered. Process damping coefficients are quantitatively estimated in a frequency domain. Then, process damping force is estimated in time domain by utilizing the process damping coefficients. A series of face turning test clarified that the turning process with process damping can be estimated accurately by using the proposed model.

Directed energy deposition (DED), which is one of the metal-powder additive manufacturing processes, is a promising method suitable for repairing. Because the carrier gas that transports the metal powder and the shield gas that protects the optical system are supplied in DED, it is difficult to estimate the transport of metal powder due to the complex turbulence around the processing point, resulting the powder supply efficiency gets lowered. This study computationally analyzed the fluid dynamics of two types of DED nozzles for improving the powder supply efficiency by changing the number of powder channels, working distance and channel diameter.

A new device for straightness measurement which integrate 10 cantilever displacement sensors has been proposed. This device was designed to measure the waviness smaller than 100 μm in height. Ten cantilevers of 11mm length were fabricated in parallel with 1.8mm pitch in a side of the base substrate of 20 mm square. The strain, induced by a displacement of the probe placed near the front edge of the cantilever, is detected as a change in the resistance of the piezo resistor at the root of the cantilever. A prototype was investigated and it was found that the resistance value increases in proportion to the displacement of the probe. Hence, it was confirmed that this device has the basic ability of displacement detection.

This paper describes an automation method applied for the generation of CNC programs. In this method, an RPA software was used to record the actions required to generate a CNC code, which later were combined to reduce repetitive actions and to avoid human errors involved in the traditional flow followed to generate CNC programs. To achieve this, an interface containing a database that automatically provides the characteristics of the workpiece, tools and cutting parameters involved on each manufacturing process was also created. This method was successfully applied in the generation of the CNC programs required to the fabrication of a standard part in five different materials, two workpiece sizes and in 3 machining centers.

Digitalization of manufacturing systems and their components has been studied for the improvement of quality, productivity, and inter-operability in manufacturing processes. In comparison, less attention has been paid to digitalization of physical actions and mental decisions of operators, engineers, and managers engaged in engineering processes. In this report, the authors analyze information flows occurred across a variety of engineering processes related to operations of machine tools in a smart manufacturing testbed by developing process (activity)-centric model following IDEF0. The model is useful to describe the overview of information flow and mechanisms (such as machine tools, humans, and engineering software tools) necessary for specific activities. The report discusses some limitations of IDEF0-based method regarding interactions among these mechanisms, and dependency of activities on these mechanisms used.

In the case of water-soluble cutting fluid, the higher concentration emulsion caused the smaller wear width VB and larger VB"'. In the case of oily cutting fluid, the lower co-efficient of friction oil caused the smaller wear width VB and larger VB"'. In both cases of water soluble and oily cutting fluid, the cutting fluid of lower coefficient of friction caused the larger wear width VB. This tendency agreed with the wear width VB"' in turning. In the both cases of water soluble and oily cutting fluid, the cutting fluid of lower coefficient of friction caused the larger cutting force volatility.

We are proposing Digital Triplet (D3) to support manufacturing system engineers in executing engineering processes on shop floors, including Kaizen activities, by fully utilizing Digital Twin. The goal of our research is to verify the feasibility and advantages of D3. For this purpose, this paper aims to develop a learning factory to embody D3. First, this paper describes the requirements of the learning factory. Next, we illustrate how to implement the factory and the result of the implementation. Finally, this paper discusses the difficulties we faced in the development.

In recent years, the industrial application of deep learning to visual inspection has been progressing along with the development of image recognition research. However, we can point out three problems with the application of deep learning to visual inspection. First, computer resources are limited when considering embedded systems. Second, tact time should be short for in situ inspection in mass-production processes. Third, high accuracy is required because overlooking a defective product can be fatal. In this study, we propose a method to build a highly accurate, lightweight, and fast inference deep learning model using a compression method called “filter-level pruning”.

Gear unit must be improved load capability and reduced its vibration and noise. In the conventional tooth surface modification design method, it was difficult to simultaneously improve the load capability and reduce vibration and noise. In this study, we made this possible by applying a shape representation method using principal component analysis to tooth surface modification design. This report describes an overview of the method and the results of the effectiveness verification by test.

We developed a new polycrystalline diamond (PCD) spherical rotary tool to achieve a mirror-finished surface on the small mold core and cavities with the surface roughness within ten nanometers Ra. In the case of the study on the mirror finishing with this PCD spherical rotary tool, we found that the actual cutting depth was about 2/5 less shallow than that the set depth of cut because this series of tools is generally small in diameter having a radius of 0.5 mm or less. In order to guarantee the dimensional accuracy of the cut surface, it is necessary to know the actual depth of cut. Therefore, we consider predicting this actual depth of cut from the processing conditions. we experimented on the set depth of cut and the actual depth of cut by changing the machining conditions within the practical range, and conducted a study aimed at constructing this empirical formula. The experiment is performed on two different types of materials that are widely used for mold cavities, the hardened stainless steel AISI420 grade modified commercially called STAVAX and cemented carbide D3 grade to determine the correlation coefficients of these materials. As results, the empirical formula under actual machining conditions was confirmed that the actual depth cut and the experimental value were in a good agreement within ±0.2 μm. It was clarified that the proposed empirical formula is practical.