Published papers

Polydimethylsiloxane (PDMS) is one of the important materials for microfluidic chips. The pattern of micro channels on the PDMS plate is usually fabricated through the photolithography and micro molding process. However, the photolithographic method requires multi chemical and mechanical processes and resultant long process time. The micro milling process is a feasible method for rapid fabrication of various patterns of micro channels. However, micromachining has not yet been applied to soft polymer materials. It is difficult to machine elastic materials such as PDMS because of their low toughness. In order to machine a micro grooves on soft polymer materials, the cutting process applying cryogenic cooling is proposed because the elastic properties of soft polymer materials remarkably change from rubbery state to glassy state below the glass transition temperature. In this study, the freezing milling method using liquid nitrogen is applied to the micro grooving of PDMS. The result of a cutting test shows that micro grooves can be shaped easily and machined accurately in PDMS by the proposed method.

Currently, viscoelastic polymers are used to coat automotive body to prevent scratches. On the other hand, the application of a polymer with high viscoelasticity makes the repair polishing process of the coated body more difficult. Hence, performing the polishing process requires the operator to be technically skilled. However, the number of workers with necessary technical skill has been decreasing because of an aging population. In addition, the surface quality and process time are dependent on the proficiency level of the worker. To automate the repair polishing process, in our past research, a serial-parallel mechanism polishing machine was developed to simultaneously control the tool trajectory, tool posture, and polishing force. The present study aims to construct a replication system of a skilled polishing technique on the basis of the above three physical parameters by applying the developed polishing machine. First, the tool trajectory, tool posture, and polishing force of a skilled worker, which reflect the polishing technique, were acquired using a high-speed camera and a dynamometer. By inputting the acquired data to the developed machine, the machine can copy the polishing technique because the tool trajectory, tool posture, and polishing force can be independently controlled. From the results of the polishing test, the surface quality after polishing was comparable to that of the skilled worker output, which satisfies the criteria of surface quality.

Actuators with clutch mechanisms have advantages in safety, backdriveability, and controllability for human-coexistence robots. This paper presents a large force actuator using electro-adhesive gels, which are functional materials controlled by the magnitude of an electric field. The developed actuator can generate and continuously control a force over 560 N and resist with a small force when de-energized. The number of electro-adhesive gel disks is drastically increased to increase the generative force. This paper reports the structural plan for the new actuator and experiments to investigate its basic characteristics. Based on the experimental results, an actuator model is identified for the future design of its control system. This paper contributes to the development of a clutch-type linear actuator with a large generative force, maintaining safety and a high force-to-weight ratio. It also addresses the problem of high-voltage power supply costs for electro-rheological and electro-adhesive materials.

The requirement for Mass production of Microlens array (MLA) mold is increasing. MLA molds are manufactured by 5-axis ultra-precision machining center using an ultra-small diameter end mill. Since visual examination is not good at judging quality of product while machining, the development of effective process monitoring technology is required. The ultra-precision machining center used in this study employs a linear motor and has low sliding resistance while machining. Therefore, machining information is directly reflected in servo information. Focusing on this point, we developed a machine learning-based shape error estimation model using servomotor current alone and discussed its usefulness.

The demand for the mass production of micro-lens arrays (MLAs) is increasing. An MLA is fabricated through an injection molding process, and its mold is manufactured by a five-axis high-precision machine tool using a small diameter endmill. A visual examination is not available to judge the quality of the mold while machining. Therefore, an effective process monitoring technology must be developed. A promising approach is to apply a servomotor current to in-process monitoring because as long as the servomotor works well, no external sensors, capital investment, or maintenance processes are required. From this perspective, a machine learning-based shape error estimation method using only the servomotor current is proposed. To explore the relationship between the motor current generated during micro-milling and the shape error of the mold, the servomotor current in X-, Y-, and Z-axes was recorded, and the corresponding shape error of the MLA mold was measured after machining. Input data were prepared by converting time-domain servomotor current data to frequency-domain data using short-time Fourier transform and reducing the dimensions of the data via principal component analysis. In terms of a meaningful label for the output data, the average shape error in the machined area corresponding to each window was provided. The input/output relationships were used to train five different machine learning models, and the accuracy of shape error estimation using each model was evaluated. In addition, the estimation accuracies using the X-, Y-, and Z-axes were compared to find the axis that senses the shape error with the highest accuracy. The results show that the non-linear method using the X-axis servomotor current information closest to the machining point achieved the highest shape error estimation accuracy.

The application of CFRP to machine tool structures can potentially improve their dynamic characteristics and energy efficiency owing to its highly specific material properties. The different thermal characteristics of CFRP and steel in spindle units likely induce unique thermal deformation, which affects the bearing dynamics. This research proposes a modeling method of CFRP spindle unit dynamics via generalized multipoint receptance coupling with contact characteristics at the spindle-holder interface. The temperature-dependent mechanical properties are analyzed by parameter identification from the model. The translational stiffness and damping of bearings in the CFRP spindle increase with temperature increase at high rotational speed.

In recent years, metal 3D printers have been attracting attention not only for building solid parts but also for repairing and coating processes. The goal of this study is to apply the directed energy deposition (DED), which is one of the additive manufacturing, to realize high-quality metal coating by depositing the coating material on a cylindrical object. To evaluate the effect of process parameters, experimental points were set up based on the central composite design and deposition was performed using Inconel 625 as the material powder. The hardness of the coating was measured and statistically analyzed by the response surface methodology. From the analysis results, it was found that laser power has the largest influence on the coating hardness, which was improved by 28% compared to the low rotation speed by depositing under the optimum processing parameters.

Polishing is an important process performed in the finishing and repair processes of mechanical parts and is still a manual work of skilled workers. However, in addition to the time and cost required for manual polishing, there are also problems such as a decrease in the number of workers and health hazards due to dust generated during polishing. From these problems, the demand for automatic polishing is increasing at manufacturing sites. To automate polishing process, a macro-micro system based polishing robot which consists of a functional end-effector and industrial robot has been proposed. Regarding end effectors, mechanical design aiming for high rigidity and motion control methods for keeping the contact force constant is being researched. Meanwhile, the tool rotation speed of the spindle is independently controlled typically. Therefore, in this research, a functional polishing module capable of polishing force control and tool rotation control according to the tool feed rate is developed as the end effector. For the polishing automation, the macro-micro system based polishing robot is constructed attaching the polishing module to the industrial robot. From the viewpoint of Preston’s law, the effectiveness of the robot polishing is investigated through polishing experiments and measuring the shape of the polished surface.

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.

There is a strong demand for automating the rough polishing process in plastic-mold manufacturing. In this study, with the active end effector being capable of fast force control and dynamic spindle control, a robot polishing control system based on a macro-micro-mechanism was built, which controls all parameters related to Preston's law and can be linked with computer-aided manufacturing. Based on the extended Preston equation, a polishing control algorithm that achieves uniform surface quality and uniform removal depth was proposed. The spherical rough polishing test on carbon steel demonstrated the efficacy of the proposed robot polishing control.

Directed energy deposition (DED) is an additive manufacturing process wherein a metal powder is fed through a nozzle onto a baseplate, melted, and solidified using a laser. Potential applications of DED have been actively discussed. In this study, laser-based DED (DED-LB) was applied as a high-speed metal-coating technique to a cylindrical pipe. The mechanical properties of DED-fabricated objects were strongly affected by processing parameters such as laser power, powder feed rate, and rotational speed of the cylindrical pipe when using 20MnCr5 and M2 as the coating materials. Hardness measurements showed that the hardness of 20MnCr5 increased with a high heat input, whereas that of M2 increased with a low heat input. The two metals exhibited opposite trends in terms of hardness variation because of the differences in their transformation points. This study analyzed a layer coated with DED-LB from the viewpoint of metal morphology.

RISA grinding method using cerium oxide slurry has been proposed for grinding optical glass lenses. While the surface quality can be maintained at high feed rate and depth of cut, there is a problem that the profile error increases toward the center of the workpiece. The workpiece is SSG and the grinding wheel is a resin diamond wheel. In this study, the relationship between dwell time and depth of cut is experimentally investigated, and dwell time control method is proposed to achieve a uniform depth of cut. This method changes the rotation speed of workpiece to keep the slurry dwell time constant regardless of the distance from the center of workpiece. The effectiveness of this method was evaluated through orbicular machining tests. With conventional RISA grinding, the grinding trace becomes deeper as it gets closer to the center of the workpiece. By using dwell time control, the depth of the grinding trace is almost the same regardless of the distance from the center.

The evaluation technique of gas permeable characterization has been developed for an increased efficiency of gas-liquid chemical reactions and high accuracy of environmental diagnosis. This technique enables us to measure spatial distributions of velocity and dissolved gas concentration by utilizing confocal micron-resolution particle image velocimetry combined with a laser-induced fluorescence technique. Microfluidic devices with gas permeability through polymer membranes are composed of a cover glass and a polydimethylsiloxane (PDMS) chip with the ability to permeate various gases, since PDMS is an elastomeric material. In the chip, microchannels are manufactured using a cryogenic micromachining system. The gas permeation is dominated by several factors, such as the gas and liquid flow rates, the membrane thickness between the gas and liquid flow, and the surface area of the membranes. The advantage of the present device is to realize the control of gas permeability by changing the surface roughness of PDMS, because the cryogenic micromachining enables us to control the surface roughness of microchannels and an increase in roughness yields an increase in the surface area of membranes. The experiments were performed under several conditions with a change in the gas flow rate, the PDMS membrane thickness and the surface roughness, which affect the gas permeation phenomena. The spatial distributions of velocity and dissolved gas concentration were measured in the liquid flow fields. The results indicate that the velocity-vector distributions have similar patterns under all experimental conditions, while the dissolved gas concentration distributions have different patterns. It was observed that the gas permeability through PDMS membranes increased with an increase in gas flow rates and surface roughness and with a decrease in membrane thicknesses, which is in qualitative agreement with membrane theory. The important conclusion is that the proposed technique is suggested to have the possibility of evaluating the characterization of gas permeable microfluidic device through membranes.

Accurate monitoring and adaptive control to maintain optimum machining conditions are required for intelligent machine tools. Methods using additional sensors have been studied to add these functions to machine tools however, incorporating additional sensors leads to higher production costs and reductions in mechanical stiffness. In the present study, a sensorless method of monitoring cutting force that utilizes the current signal and the second-order derivative of position is developed. Moreover, applying sensorless cutting force monitoring, a hybrid control method is proposed in which the sensorless monitoring is used to simultaneously control both the position trajectory and cutting force in the feed direction.

Robots manipulated by humans are being aggressively researched and developed. Such robots have requirements such as a large generative force, high-speed response, good controllability, high safety, low friction and backdriveability. We focused on backdriveability and created a prototype for a novel linear actuator with a new structure that includes an electrorheological (ER) gel, which is a recently designed functional material. The device is an all-purpose actuator that satisfies the demands for application in a human-coexistence welfare robot, such as power-assist systems. This paper presents the original plan for the linear actuator, shows the actuator as currently manufactured and presents the basic characteristics of the actuator as investigated experimentally. We obtained larger force from the linear actuator than the initial target. Additionally, the linear actuator shows possibilities of further improvement of the force-weight ratio by an increase of the ER gel structures. With regard to stabilizing the response of the linear actuator, spring elements contributed to its improvement. However, further stabilization remains as a future issue for practical use in human-coexistence robots. (C) Koninklijke Brill NV, Leiden and The Robotics Society of Japan, 2010

Viscoelastic polymers are used as one of coating materials for protecting the products from scratches. Presently, the repair of the coating surface for removing dust or extraneous matters has been performed through several polishing processes. However, it becomes increasingly difficult to polish its surface by only applying good skill and experience of skilled worker because a leading-edge viscoelastic polymer for coating is further scratch-resistant. Thus, based on quantitative evaluation of relation between polishing process and finished surface, it is necessary to make the polishing process appropriate for the leading-edge viscoelastic polymers. In this study, we attempt to establish the evaluation method of the polished surface and clarify the surface condition with invisible scratches.