Weldline that is formed when two or more melt fronts meet is one of the major defects in plastic injection molding (PIM), which has an influence on not only the appearance but also the strength of a plastic product. Consequently, it is preferable to reduce it as much as possible for high product quality. On the other hand, high productivity is always required in the PIM. In other words, cycle time should be minimized. To achieve the high product quality and productivity, process parameters such as melt temperature, cooling time and so on plays an important role. In this paper, the process parameters in PIM are optimized for minimizing the weldline and the cycle time simultaneously. In particular, variable injection velocity that the injection velocity varies during the PIM process is adopted and optimized. A novel weldline evaluation is also proposed. Numerical simulation in PIM is so intensive that a sequential approximate optimization (SAO) using radial basis function (RBF) network is adopted for the design optimization. It is found from the numerical result that the trade-off between the weldline reduction and the cycle time is observed. Based on the numerical result, the experiment using the PIM machine (GL100, Sodick) is carried out. Through the numerical and experimental result, the validity of the proposed approach is examined.

This study discusses the verification of the control system for realizing autonomous standing up from the parking mode of a novel motorcycle with a self-stabilizing mechanism, named “MOTOROiD,” and ensuring its stability on low-speed driving. MOTOROiD has a novel rotary axis that can vary the position of the total center of gravity, referred to as the active mass center control system (AMCES). In the previous study, the authors derived a mathematical model using Lagrange’s equation of motion, designed a control system for standing up from the parking mode, and ensured the motorcycle’s stability by using a two-degree-of-freedom control system based on the established model. However, the robustness of the proposed system was not fully verified. The previous study only verified the robustness of the mass variation with a numerical simulation example. Therefore, this study demonstrates the effectiveness of the proposed control system using several numerical simulations with variations in the system parameters. The simulation conditions are set by considering the characteristics of MOTOROiD in realistic situations related to a parked motorcycle. We consider three cases: (i) MOTOROiD is parked on a sloped ground, (ii) the angle of inclination at parking mode is not as assumed, and (iii) MOTOROiD is influenced by Coulomb friction. We set the simulation parameters based on these cases. Consequently, we confirm that MOTOROiD can stand up autonomously under all these conditions, thereby demonstrating the practicality of the proposed control system.

This study describes an approach to quantify the effect of the bending moment on the fatigue strength of a bolt in bolt/nut assembly. To confirm the validity of the conventional fatigue design methodology based on the nominal stress acting on the specified nominal stress area, both the fatigue tests with various bending stress ratios (nominal bending stress/ nominal total axial stress) and the 3D-FE stress analysis for the corresponding conditions were conducted. The results from fatigue tests using a newly developed fatigue testing fixture clearly show as bending stress ratio increases, the (virtual) fatigue strength expressed by the nominal stress also increases. The results from 3D-FE analysis show that the magnitude of the local bending stress on the thread root is lower than the one estimated from nominal axial stress and the stress concentration factor for purely tensile loading. The results, however, also show that the magnitude of the local stress acting on the bolt thread root is affected by the angular position of nut against the loading axis due to the existence of the incomplete thread of nut at the bearing face side. Finally, it is concluded that the conventional fatigue design methodology is practically acceptable for tensile force combined with bending moment loading, albeit with results that are slightly conservative.

Ionic liquids have high potential as novel lubricants because of their unique physical properties. In particular, halogen-free ionic liquids are low environmental-load lubricants. Unfortunately, these ionic liquids exhibit a poor tribological performance against bearing steel. However, they exhibit high performances against diamond-like carbon (DLC) under the high vacuum condition. The influence of DLC species on tribological performances and the reaction mechanism of halogen-free ionic liquids on the worn surface are still unknown. This investigation evaluates the tribological performances of halogen-free ionic liquids (1-ethyl-3-methylimidazolium dicyanamide ([EMIM][DCN]), 1-ethyl-3-methylimidazolium tricyanomethane ([EMIM][TCC]), and 1-ethyl-3-methylimidazolium tetracyanoborate ([EMIM][TCB])) against two types of DLC (tetrahedral amorphous carbon (ta-C) and hydrogenated amorphous carbon (a-C:H)). The tribological performances of ionic liquids differed by the anion structure and the type of DLCs. The DLCs lubricated with [EMIM][DCN] and ta-C lubricated with [EMIM][TCC] exhibited a low friction coefficient. On the other hand, other ionic liquids exhibited the high friction coefficients against DLCs. A time-of-flight secondary-ion mass spectrometer analysis indicated that the tribo-decomposition of halogen-free ionic liquids and the adsorption of anion achieve a low friction. The chemical activity of nascent surfaces plays a very important role in achieving tribo-decomposition. However, because this activity of DLC was lower than that of bearing steel, [EMIM][TCC] exhibited a poor tribological performance. Thus, although the tribological performances of halogen-free ionic liquids can be improved by using DLC as sliding materials, there is also the possibility of inhibiting the formation of an adsorption layer derived from ionic liquids.

High efficiency machining achieved by shortening preparation time has attracted attention for its many advantages in the field of machining. In order to reduce preparation time, computer-aided process planning (CAPP) is expected to automatically determine machining process parameters such as machining sequence, cutting tools, and cutting conditions. In the author’s previous study, novel machining features are proposed to derive machining process parameters from specific geometric shape patterns of removal volume during a parts machining operation. It is also confirmed that the machining features are recognized by deciding the machining sequence for each machining primitive that is first obtained from the removal volume. However, workholding devices have not been considered to determine the machining sequence or to recognize the machining features in parts machining operation. Workholding devices are generally indispensable for parts machining and are practically important to determine machining sequences. Therefore, this study aims to develop a CAPP system for parts machining considering the use of workholding devices. Two general workholding devices, a machine vise and strap clamps, are assumed for various parts machining operations. Under the limitations of workholding devices, workpiece fixing positions are calculated before the machining features are recognized through allocating the machining sequence. Moreover, the number of fixing changes and the machining time are estimated according to both the workholding device being used and the workpiece fixing positions. Case studies are conducted to confirm the usefulness of the developed CAPP system considering the use of workholding devices. The results reveal that fixing positions are obtained properly for various combinations of workpieces and workholding devices. Furthermore, it is confirmed that the CAPP system has the potential to be effective for automatically determining the machining process parameters of parts machining.

This paper discusses the manufactured one-axis controlled bearingless motor and its experimental results in liquid nitrogen. The bearingless motor is an axial type bearingless motor. It generates a force in the axial direction to control the axial displacement of the rotor and a rotation torque for the motor. The axial displacement is actively controlled using a PID controller, and the rotation speed of the rotor is also actively controlled using a PI controller. The other degrees of freedom are passively supported by the repulsive force of the permanent magnet bearings. The axial type bearingless motor consists of a stator, a rotor, three displacement sensors, and three Hall sensors. Displacement sensors and Hall sensors are used to measure the axial displacement and rotation angle of the rotor. The stator has six salient poles, and the rotor has four poles with four permanent magnets. The air gap between the stator and rotor is 1.5 mm at the center position. The axial type bearingless motor and permanent magnet bearings are set in liquid nitrogen in the experiment. The experimental results show an impulse response at 0 rpm, the relationship between the displacement of the rotor and rotation speed, the relationship between the driving current and the rotation speeds, and a step response from 500 rpm to 1,000 rpm in the rotation speed. From experimental results, it is confirmed the stable levitation and rotation of the bearingless motor at liquid nitrogen temperature.

This paper evaluates the sensitivity of a proposed crack detection method of POM (Polyoxymethylene) gears using a deep convolutional neural network. The vibration signal was collected from an automatic data acquisition system for endurance tests of gears. The fast Fourier transform (FFT) of the measured vibration signals generated grayscale images for training input. A high-speed camera captured cracks at the tooth root, and the length of cracks was computed as a damage index for training labels. A convolutional neural network (CNN), called VGG16 ConvNet, which has 1000 classes in output, firstly was pre-learned from image data of ImageNet and then the weights of two layers, which were close to the output layer, was relearned from the created images of meshing vibration data with the transfer learning technique. The output layer was modified to fit two classifications problem related to the cracked or non-cracked situation of gears. The accuracy rate for the recognition of the gear fault reached 100%. However, the remained problem is whether the performance of the developed system is susceptible to the change of the working condition of gear, such as high rotational speed and torque, or not. Hence, the robustness of the crack detection performance of the developed system was investigated. The endurance tests of gears under some test conditions, such as high-low rotational speed and/or torque, were carried out to collect the different vibration signals. The accuracy rate of gear failure classification under various working condition was judged and the factors affected on the performance of the developed system under working condition changing was discussed. The results showed that the developed system learned from one testing condition incapably perform in varied testing conditions. In other words, the developed system must be learned from diversity data for a superior effectuation. In this case, our interest is to uncover how many experiments and images for re-learning are required in each experiment for better performance. The investigation of the re-learning in this paper showed the required number of images was 200 and a single endurance test under each condition was enough if an appropriate number of images were obtained.

The main objective of this research was to stabilize the motions of a walking assistive device on the stairs. We proposed a posture-compensated method suitable for the stairs and proved the proposed theory in simulations and experiments. To adapt elderly people to daily life, an assistive device that could support the walking motion and protect the users from falls was utilized. Because the target users needed to walk on the stairs where they could easily fall, the device adjusted posture to stabilize users. Zero-moment point (ZMP) was used to quantify the stability of the device. To prevent collisions during the compensation process, the assistive device maintained the stride length because the width of the stairs is invariable. The target positions of the device feet were moved the same distance. The projection of gravity and the inertia force was kept in support polygons that varied for different walking phases to remain stable. Particularly, in simulation, the virtual slope method was quoted to calculate the location of ZMP. In experiments, pressure sensors were employed to measure ZMP. The results of the experiments showed that the assistive device, employing the proposed method, guided the user to successfully maintain the ZMP on the footplate. In conclusion, the proposed method is available for stabilizing various assistive devices, especially for slow-walking motions on stairs.

To fabricate various parts with complex shapes in the sheet-metal bending industry, the V-die bending process is commonly applied. To achieve precisely bent sheet-metal parts, the design of a proper V-bending die which could prevent the spring-back characteristic and obtain the required bend angle as well as control and maintain the required bend radius is needed. The predicted spring-back and bent radius could be basically calculated by using the spring-back factor. In the present research, therefore, the accuracy of the spring-back factor for bend angle prediction and bend radius calculation in the V-die bending process was examined based on the finite element method (FEM) and laboratory experiments. The two existing spring-back factors, including the conventional and adjusted spring-back factors, were investigated. The results showed that, compared with experimental results, the accuracy of the bend angle prediction and the bend radius calculation obtained using the adjusted spring-back factor were better than those obtained using the conventional spring-back factor. These results were clarified based on the stress distribution analysis of the sheet-metal bent parts. Therefore, the adjusted spring-back factor was recommended over the conventional spring-back factor for V-bending die design applications to achieve a better accuracy bend angle prediction and bend radius calculation.

This paper describes the relationship between the oil film behavior and cooling effect of traction drive roller experimentally. Traction drives are transmission systems where the power is transmitted by shear stress of lubrication oil on the contact point between the rollers. They have advantages regarding noise, vibration, and high-speed operation compared with gears. On the other hand, when the power is transmitted, the heat generation occurs caused by the creep and spin at contact points. This leads to decrease of traction force and then leads to reduction of mechanical efficiency of power transmission. Thus, traction drives need the additional cooling device or the supplying of a large amount of cooling oil. However, these cooling method causes an increase of device weight, device cost, and energy consumption. From this background, the authors focused on the cooling effect of oil film behavior on traction roller. In this study, the oil film behavior was observed, and oil film temperatures were measured. As a result, it is found that oil film temperatures are affected by the oil splashing from the roller surface and the oil film distribution. Further, the authors proposed a new geometry of the roller surface which can expect to increase the cooling effect by the oil film.

Cellulose nanofibers (CNFs) are synthesized by unraveling cellulose made from wood. CNFs have attracted significant attention among biomass materials owing to their excellent mechanical properties and wide range of applications. This study aims to elucidate the effects of CNFs as a lubricating additive. In order to accomplish this, CNFs were dispersed in water at densities of 0.02.1.0 mass%, and friction tests were performed using a ball-on-plate friction tester with a reciprocating motion configuration. Stainless steel (JIS-SUS304) and polyoxymethylene (POM) were used as sliding plates, while various materials were used for sliding balls. On both the SUS304 and POM plates, friction reductions were observed by adding only a small amount of CNFs (0.02 mass%). Although wear on the SUS304 plate reduced by 0.02 mass%, the water dispersion of CNFs was not distinct; it was observed only at densities of CNFs that were more than 0.2 mass%. On the POM plate, wear was induced by adding 0.02 mass% of CNF water dispersion. Adverse effects may arise with the addition of more than 0.2 mass% CNF. A distinctive feature observed in the water dispersion lubrication of CNFs was that a tribofilm is not formed due to the chemically inert nature of CNFs. It appears highly probable that in the water dispersion lubrication of CNFs, the physical effects of these fibers are much stronger than their chemical effects. These effects include rolling and/or sliding, which occurred in fullerene and carbon nanotubes.

Slider-crank mechanisms are used as a rotational-to-linear motion transformer in positive displacement pumps, because such mechanisms are effective for generating high discharge pressure. The stroke of slider-crank mechanism in a water pump is twice the crank length and the connecting rod must be sufficiently long to avoid problems caused by the side force, thereby, limiting the extent to which a positive displacement pump can be miniaturized. Small pumps with high discharge pressure are expected to be used as the power source of a hydraulic actuator. In this study, a mechanism wherein eccentric cams replace a crank and a connecting rod is proposed, and the operating principle and characteristics of this mechanism are investigated. The mechanism has a stroke that is four times the crank length, and it is designed with reference to an Oldham coupling mechanism. Replacing a crank with an eccentric cam is expected to reduce vibration by reducing fluctuations in the position of the center of gravity, but a concern is that using an eccentric cam will cause friction loss. An experimental device is manufactured to investigate the driving torque and the vibration characteristics.

This study involves experimental investigations of the hydrodynamic lubrication performance of dimpled parallel thrust bearings. Circular and rectangular dimples were created on lubricating surfaces and the hydrodynamic effects were evaluated. The load carrying capacity and the frictional torque were measured at a constant film thickness. The lubricating surface was observed to make clear the effect of the cavitation area on the load carrying capacity and the frictional torque. The circular dimple produced the highest load carrying capacity in all experimental conditions tested. The frictional torque for all dimpled specimens was lower than that of the dimple-free specimen especially at higher rotational speed. At lower rotational speed, cavitation did not appear and also the load carrying capacity did not generate. Cavitation emerged at the leading edge of each dimple as the rotational speed increased, and the growth of its area was dependent on the shape of dimples. The cavitation size of the circular dimple remained constant at the rotational speeds used in the tests, whereas those of the rectangular dimples linearly increased.

In this study, the tool wear mechanism of a TiN coated carbide insert in up-cut end milling of AISI 1050 at different feed rates, was investigated. The diameter of the end mill holder that was used was 35 mm. The edge roundness of the used cutting insert was approximately 4 μm. The feed rate was varied from 2.5 μm/tooth to 150 μm/tooth at the cutting speed of 100–150 m/min. The axial depth of the cut was set at 1.0 mm and the radial depth of the cut at 5.13 and 17.5 mm. In dry conditions, the flank wear width, at the same real cutting length, increased rapidly with the feed rate and reached its maximum at 10 μm/tooth. When the feed rate was set at smaller than 10μm/tooth, the smallest tool wear occurred at the lowest cutting speed. Subsequently, the wear width decreased drastically to one third of the maximum. At this feed rate range, the larger wear width occurred at the lower cutting speed. With further increase of the feed rate, the wear width increased with the feed rate. The feed rate, at which the wear width started to increase, was lower at higher cutting speeds. The deference of tool wear in the radial depth of the cut, at the same number of cuts, increased with a feed rate smaller than 10 μm/tooth. When the feed rate exceeded 10 μm/tooth, the difference decreased according to the feed rate.