Although glitches due to motion error of the feed drive systems are typically observed on the machined surface, the relationships between the motion error and glitches are not investigated up to now. The purpose of this study is to clarify the influence of motion error of feed drive systems onto machined surface generated by ball end-mill. In order to achieve the purpose, cutting tests and simulation of hemisphere shape are carried out. The cutting tests of hemisphere are carried out with two kinds of tool paths; uni-directional and bi-directional scanning paths. Parameters of the friction compensator of the NC controller are also changed to investigate the influence of motion characteristics of the machine. In addition, a simulation method for machined surface with the dynamic model of feed drive systems is newly developed. As the results of the cutting tests and simulations, it is confirmed that the proposed simulation method can accurately predict the influences of the motion errors and tool paths onto the machined surface generated by a ball end mill. It is also clarified that the both of motion path and motion errors influence the machined surface, because the innermost motion paths to the work piece are copied onto the finished surface. For example, even if a stepwise error exists on the motion trajectories, the error is not copied onto the finished surface when the surface is generated by bi-directional tool paths, although the error clearly copied on to the surface generated by uni-directional tool path.
Edge imperfections are often introduced on workpieces due to plastic deformation during machining. These imperfections are known as burrs. Since the deburring operation is costly, the control of burr formation is a research topic of great significance for industrial applications. This study introduces a system that focuses on the prediction of burr dimensions and positions in the end milling process as a preventive measurement method. This system is based on burr formation models, analytical cutting force model, and experimental validation. Two kinds of burr models were used, rollover burr and Poisson burr. The orthogonal and oblique cutting were also used in the system based on different positions. A Window based program is developed to illustrate the machining process upon PC-based NC simulator that consisted of geometric simulator and physical simulator. The geometric simulator consists of the feature identification and the cutting condition identification. The physical simulator contains cutting force model that used to calculate the force in feed direction that leaded burr to form. The propose system was compared with experimental in difference workpiece materials for validation. It was verified that top burr and exit burr in up milling and down milling can be predicted. Both the predicted and experimental results were found to agree in most of the cutting conditions.
The surface roughness is one of the most important characteristics of machined parts. A systematic method is proposed in the paper to simulate the shape generation processes in the boring operations and to estimate the surface roughness of the generated faces, based on the machining parameters. The simulation model includes both the models of the shape generation motions considering kinematic motion deviations and the cutting tool geometries. The shape generation motions with deviations are mathematically described by 4 by 4 transformation matrices. A set of points on the bored holes are generated through the simulations, and an assessment surface is obtained as the datum reference to estimate the 2-dimensional (2D) and the 3-dimensional (3D) surface roughness, based on the points generated by the boring process simulations. The proposed method provides us with a systematic method to estimate the surface roughness in the boring processes including the kinematic motion deviations.
A method to generate a high-continuity thickness measuring path has been proposed in our own developed automatic system using laser displacement sensors and an industrial robot to measure and evaluate a free curved plate thickness. Besides the system generates the thickness measuring path on the basis of the point cloud data representing a workpiece shape, in this method, a quaternion is used to represent a change in the measuring posture with squad which is an interpolation algorithm. This paper presents a new method to determine the positions and the postures of the representative points for continuous interpolation. These positions and the postures are determined using 3D Configuration space(C-Space) consisting three parameters for representing the measuring posture. For details, the measurable postures are mapped in the C-Space on each measuring point. Then, the positions of the representative points are determined using a local maximum value of the distance between centroid coordinate of the measurable postures in adjacent points. Moreover, the postures of them are determined using the common postures of adjacent points or the distance from the origin of the C-Space. The continuity of the measuring path is evaluated from the change in the joint angle of the robot. Specifically, angular acceleration of the joint angle is used to evaluate for stabilization of the high-speed robot motion. The experimental result shows that the angular acceleration of the robot joint in generated path is smaller than that of the path generated by a conventional method.
Ceramics materials have excellent mechanical characteristics such as high-stiffness, low-thermal expansion, wear resistance and light weight. Owing to those characteristics, the ceramics parts have been used for the sliding part of the precision positioning stages. However, it is difficult for the ceramics workpieces to achieve the desired form accuracy in once machining process because of the brittleness of the ceramics materials. The compensation of the form error is processed by the manual procedures in general, so it requires a skill of the operator and a great deal of time. In order to realize the automation of the error compensation, it is necessary to measure the workpiece form on the machine tool. In this study, an on-machine form measurement system using a laser displacement sensor having nanometric accuracy is introduced for the surface form measurement of the precision ceramics workpieces. The laser displacement sensor is mounted on the tool post of the ultra-precision polishing machine to realize the on-machine measurement of the ceramics workpieces. By using two horizontal linear slides of the machine tool, the laser displacement sensor can scan the surface of the workpiece with sub-micrometric motion accuracy. Consequently, three-dimensional continuous surface form of the precisely machined ceramics workpiece can be obtained in non-contact condition. In order to evaluate the measurement accuracy of the constructed system, the experiments to investigate stability and motion accuracy were conducted, and then the form measurement of the precision ceramics parts is carried out by using the proposed measurement system.
We propose a measuring technique for microflow based on the stochastic resonance phenomenon. The minute changed signal from the microflow hidden within the threshold of multi-threshold imaging sensors, such as CCDs, can be reconstructed using an external random signal. The differences in optical path length fringes of a Michelson interferometer were enhanced by scattered light using a micro Brownian particle solution. In this paper, we investigate the required characteristics of the external random signal numerically and experimentally. The feasibility of the proposed method was experimentally confirmed. The number of pixels required to reconstruct the signal was smaller than expected owing to internal and environmental noise signals. We show that the standard deviation of the external random signal plays an important role in enhancing the accuracy of the measurement.
This paper presents a method for the measurement of the contact potential difference (CPD) and material distribution on the sample surface by using a scanning electrostatic force microscope (SEFM). In SEFM, the surface profile is evaluated by applying an electrostatic force generated between the probe tip and the sample surface. The amount of electrostatic force relies on both the surface profile and the surface material distribution. To selectively extract the quantitative surface profile, the calculation method to cancel the effect of surface material distribution towards detected electrostatic force is applied in the profile measurement using SEFM. On the other hand, by utilizing the variation of electrostatic force affected by the material distribution, and applying the quantitative profile measurement principle, CPD measurement can be realized. Since the intensity of electrostatic force generated between the probe and the sample surface will rely on the differences of the CPD, the CPD can be calculated by using the frequency shifts of the probe oscillation when two different bias voltages are applied between the tip and the sample. In this paper, firstly, the effect of the surface material distribution on the profile measurement result is reported. The profile measurement result of a sample which consists of two or more materials showed a similar profile with the measurement result obtained by a commercial AFM, which demonstrates that the proposed principle is effective in quantitative measurement of surface profile. Then, the detection sensitivity of the CPD corresponding to the experimental conditions has been confirmed by simulation. In addition, the basic characteristic of the CPD measurement system has been evaluated.
A new scan mode referred to as the feed-forward controlled unidirectional scan mode is employed to improve the accuracy of surface profile measurement of micro-optics in a scanning electrostatic force microscope (SEFM) instead of the conventional feedback controlled bidirectional scan mode. Two kinds of error sources, the positioning deviation of the X-stage and the frequency shift detection delay due to the lock time of the PLL circuit, are analyzed theoretically. It is verified that large profile measurement error can be introduced by the error sources when steep structures are included in the surface profile. It is also demonstrated that the feed-forward controlled unidirectional scan mode can remove the influence of the positioning deviation error component. It can also reduce the influence of the frequency shift detection delay error component by shortening the lock time of the PLL circuit and increasing the bandwidth of the PLL circuit through a feed-forward control strategy of the scan traces with the PI controller of the Z-scanner turned off. Experiments of surface profile measurement are carried out on two diffraction grating samples with different grating structures. The feasibility of the new scan mode is confirmed by the experimental results.
A fabrication process of a contact detection sensor, which is thermally-sensitive thin film sensor referred to as a thermal contact sensor to be used for a surface defect inspection, is designed based on photolithography process. The existence of a small defect, the size of which is on the order of several-ten nanometers, would be detected by capturing frictional heat generated at the contact between the thermal contact sensor and the defect. The thermal contact sensor is therefore designed to detect small quantity of frictional heat, the rate of heat supply of which is estimated to be on the order of microwatts. In this paper, followed by the fabrication of the first prototype thermal contact sensor in the previous study, the structure of the thermal contact sensor is modified to make the thermally-sensitive area of the thermal contact sensor to be located at its top surface. For further stable fabrication of a thermal contact sensor, lift-off process with a positive resist is also introduced in the sensor fabrication process. Experiments are carried out to verify the feasibility of the fabricated second prototype thermal contact sensor. In addition, aiming to improve sensitivity of the thermal contact sensor, the relationship between the sensor sensitivity and the sensor bias voltage is investigated in experiments. Furthermore, FEM simulation is carried out to estimate the increase of the sensor temperature due to a self-heating effect by the bias voltage applied to the thermal contact sensor, since the sensor temperature deviation could affect the sensitivity of the thermal contact sensor.
An on-machine measurement system to measure the workpiece's geometry or dimension plays an important role in manufacturing processes. A laser displacement sensor recently attracts more attention which enables on-machine high-speed continuous scanning measurement. We developed an on-machine measurement system using a triangulation-based laser displacement sensor for five-axis machine tools. By driving the machine's rotary axes to tilt or rotate the workpiece, the object's geometry can be measured from all the directions in the three-dimensional space without being detached from a machine table. In such a measurement, location errors and error motions of the machine's rotary axes clearly become major contributors for the measurement uncertainty. To identify location errors of rotary axes by using the on-machine laser measurement system itself, this paper presents its application to the R-test, where the three-dimensional displacement of a precision sphere, attached to a rotary table, is measured in the machine tool coordinate system. In the present laser measurement system, the position and the orientation of the laser displacement sensor must be calibrated in advance by using the motion of the machine's rotary axes. The location errors of rotary axes cause their identification error. Unlike the conventional R-test, where the sphere's displacement is measured by three contact-type displacement sensors, the formulation to identify location errors must consider the influence of this calibration error. Experimental demonstration is presented.
Demands for nanometer positioning with a high resolution and a long stroke have increased in a variety of industries. Performance of guide elements is one of the most important issues for realizing such positioning systems. In general, an ultra-precision positioning system has aerostatic guideway which can restrain a moving table without non-linear behavior, i.e., friction and backlash. Characteristics of aerostatic guideway affect the performance of the positioning system. This study is aiming at developing an aerostatic guideway which can achieve a high speed nanometer positioning. The aerostatic guideway preloaded by magnetic attraction force can lead to provide a compact structure, and the characteristics can be easily controlled without increased driving force. However, the relative velocity between permanent magnets and the attracted surface causes the induced eddy current, which deteriorates the performance during high speed positioning. In this study, in order to minimize the eddy current loss, the magnetic attraction force is successfully added to a mechanism following the positioning table. In order to evaluate the proposed structure, a positioning system is developed by using the magnetic attraction force preloaded aerostatic guideway. The positioning system is constructed by a positioning table and a following table. Basic characteristics of the aerostatic guideway are experimentally evaluated. From the results, the magnetic attraction force-preloaded guideway can be designed so as to obtain high stiffness. By optimizing the magnetic force, the performance of the positioning system was evaluated. The experimental results of frequency response confirm that the response of the positioning system can be improved by applying magnetic attraction force. In addition, the results of stepwise positioning confirm that the proposed positioning system has high positioning resolution. These results confirm that the proposed guideway achieves high speed nanometer positioning capability.
The development of nickel mold materials that softened at high temperatures was investigated, and their use in the fabrication of replica nickel molds by thermal nanoimprinting was demonstrated. Ni-based carbon nanotube (CNT) composite coatings were formed by ultrasonic assisted electroplating using a horn sonicator. 1.6 g/L CNTs, which were 9.5 nm in diameter and 1.5 μm in average length, were added to a nickel sulfamate plating bath. The Vickers hardness of the Ni-based CNT composite coatings was over 500 HV at room temperature and under 50 HV at high temperatures over the range of 400 to 600 °C. After heat treatment at 500 °C, the grain size of the Ni-based CNT coatings became larger than that of normal Ni coatings. Ni-based CNT composite coatings had random crystals, instead of the columnar crystals in pure nickel coatings. It is thought that the softening of the Ni-based CNT composite was based on containing CNTs in grain boundaries, as well as a crystal structure transformation. To fabricate the Ni-based replica mold, a Ni-based CNT composite coating was directly imprinted at 3 MPa and 500 °C; higher than the softening temperature of the coatings. 5 μm square dot patterns from a mother silicon mold was successfully replicated by thermal nanoimprinting on the Ni-based CNT composite coating. The fabrication method of replica nickel molds was based on the high-temperature softening property of Ni-based CNT composite coatings.
This paper defines static and dynamic component parameters based on the method that converts thrust and torque detected during drilling process into equivalent thrust force and principal force. Features of the parameters are extracted by wavelet packet transform (WPT) and then used to train a back propagation neural network (BPNN) to predict the drill wear. Experiments with different drilling conditions and workpiece materials were conducted and it has been confirmed that both static and dynamic component parameters are affected by the drilling conditions. The features extracted from dynamic components in lower frequency band can predict the drill corner wear better.
The standard of test method for five-axis machining center is now under discussion by ISO. In its draft standard, the use of a measurement device with three displacement sensor is described in addition to the ball bar device that is well-known as the motion measurement device for machine tools. In this study, a device that was able to take the setting method freely was designed and prototyped. Using the newly-designed device, three setting method installed on the machine were compared. One is of the conventional setting method developed in Europe, another is the method that sensors axes are parallel to the machine axes, the other is that the device is attached on the machine spindle. Since the compared results show little difference, the setting method can be decided by convenience of the easiness of installation on the machine such as centering of the master ball, collision prevention and sensor's cable treatment. The measured results on two five-axis machining centers were also considered. The typical error was that the straightness error motion of linear axes, which is difficult to detect by the ball bar test. Both bidirectional repeatability error and cyclic errors are detected as the straightness error.
This research aims to develop an automatic process planning system based on the machining feature recognition in the complex machining for Turning-Milling Center. The previous studies on the machining feature recognition are briefly discussed. In this study, the machining feature recognition is conducted based on the delta volume decomposition to achieve optimal result of process planning. The complex delta volume is cut or disassembled to generate simple machining features. Each surface of the delta volume is the candidate of the cutting plane. By this disassembly method, various possible candidates of machining features for process planning are obtained from the delta volume. The Solidworks API has been used in the designed system for automatic disassembly of the delta volume into simple machining features, feature recognition, tool-path length calculation, sequence determination, process selection and prediction of processing time. In this research, a new approach to machining feature recognition has been developed based on a design table and surface comparison method. Further analysis to select the best candidate of machining features are done automatically by applying several machining rules. This process planning system is able to evaluate all possible machining solutions and sequences, and determine the machining plan which has the shortest machining time.
We have developed a new machining method which realizes turning of non-axisymmetric curved surfaces with a rotation axis (called NACS-Turning hereafter). NACS-Turning is a CNC turning method with 3-axis synchronous which is composed of a turning axis and two translation axes. This method forms a profile of a non-axisymmetric curved surface by adopting a liner motor in X axis and by synchronizing the X axis with the rotation axis at high speed. We confirmed dramatic improvement of productivity by using this machining method. However, at present, the generation of tool paths for shape forming using NACS-Turning are obtained with APIs associated with a commercial 3D-CAD program and geometric calculations. Therefore, we have developed a new CAM application for NACS-Turning in this paper. Since this machining method is three-axis simultaneous control, even though composition of the axis is the same as a general two-axis turning lathe, tool paths have to be generated in three dimensional space. Therefore, applying for a tool generation method of a machining center is required. Specifically, machining points are generated by calculating intersection lines on free surfaces as machining surfaces are calculated, and coordinates of tool center points are calculated by using the inverse offset method. In this paper, first, outline of NACS-Turning is mentioned. Next, the detail of the developed CAM system which is limited to a rotary tool. Finally, we report effectiveness by cutting experimentation.
The objectives of the present research are to investigate relationships between kinematic motion deviations of machine tools and geometric tolerances of their components, to propose mathematical models representing kinematic motion deviations of five-axis machining centers, and to apply the models to analysis of the kinematic motion deviations of five-axis machining centers. A set of models is proposed to represent kinematic motion deviations of both the linear tables and the rotary tables, based on the geometric tolerances of the guide-ways connecting the components. By combing the models, three models of the five-axis machining centers are developed and applied to the analysis of the standard deviations of the shape generation motions of the tools and the workpieces. The proposed models provides us with a systematic method to analyze and to estimate the kinematic motion deviations of the five-axis machining centers, based on the geometric tolerances of the guide-ways connecting the components.
The goal of the production-planning problem is to find the optimal solution from combinations that satisfy constraints. Traditionally, production-planning problems have been treated separately, due to the problem size and complex constraints, whereas the constraints in production planning spread to the entirety of production planning. For example, when the dynamic changes as machine breakdowns and new machine addition occur, a production plan should be modified with new constraints caused by these dynamic changes. Therefore, it is required to represent a set of solution candidates that satisfy constraints in production planning. Zero-Suppressed Binary Decision Diagram (ZDD) is a directed graph representation of Boolean function and can efficiently represent a set of combinations. This paper describes the validity of application of the ZDD to production planning to represent a set of solution candidates. Experimental results applied to a sample-planning problem demonstrate that the ZDD is efficiently used for production planning and for dealing with dynamic changes.
A cellular manufacturing system is a kind of lean and flexible manufacturing systems in which a worker or a group of workers carry out all assembly processes of a product. Assembly times of workers are generally improved in an exponential manner by repeating same assembly processes. However, the workers in the cellular manufacturing system are not provided with enough time to learn the assembly processes, since product mixes and production volumes are changeable in a short period of time. The objective of this study is to propose a work instruction system for untrained workers in assembly cells to understand the assembly processes quickly and reduce their assembly times of products without repeating same assembly processes. This study firstly records workers' assembly processes by using video equipment and analyzes the learning processes of the workers in order to propose an effective strategy for the workers to reduce the assembly processes in the cellular manufacturing system. According to the proposed strategy, a prototype of a work instruction system is developed for untrained workers in assembly cells. The prototype system provides graphical user interfaces explaining the information of assembly processes for the workers to facilitate understanding the assembly processes. Some experiments are carried out for assembling same toy cars built with Lego blocks in order to evaluate the effectiveness of the developed work instruction system. Experimental results of the proposed work instruction system are compared with the ones of a simple work instruction system. Through the comparison, it is recognized that the proposed work instruction system is superior to the simple work instruction system from the viewpoint of the reduction of assembly times. At the last part of this study, we propose a prediction method of assembly times of workers by measuring biological information as a heart rate, in order to carry out dynamic production management through the work instruction system.
This paper proposes a non-contact and on-machine measurement method for evaluating tool edge geometry. In the proposed method, a focused laser beam having a diameter of several micrometres traces over the tool edge. By utilizing a light intensity of the laser beam passed around the tool rake face, the gap between the centre of the optical axis of the focused laser beam and the tool edge can be obtained. By combining the measured gap and the information on the XY positions of the focused laser beam, the tool edge contour can be evaluated. In the proposed method, stability of the laser power emitted from a light source would affect measurement accuracy of the tool edge contour. A modified optical design was therefore applied to the evaluation system so that a laser power drift and influences of common-mode noise could be compensated in real time. A modified evaluation system consisting of a laser diode, a beam splitter, a pair of lenses and two photodiodes was developed. Experiments were carried out to test the basic performances of the developed evaluation system with the modified optical design. Possible sources of measurement errors in the tool edge contour evaluation were also discussed. Furthermore, computer simulation was carried out to confirm measurement resolution of the developed system along the tool edge contour.