Titanium aluminide bonded diamond wheels based on aluminum alloy were fabricated by centrifugal casting method. Main features of the wheels include lightweight aluminum alloy for the wheel-base, the titanium aluminide compounds as a metal bond matrix, and the monolithic construction-wheel. Basic and practical grinding tests were carried out to confirm the compatibility of the aluminum alloy containing titanium aluminide phase and the grinding performance. Two different kinds of workpiece, alumina ceramics and cemented carbide, were chosen for use in these tests. The results obtained are as follows : (1) Mechanical properties, hardness, wear resistance, bending strength, of the aluminum alloy containing the titanium aluminide phase suitable for the bond matrix of the wheel. (2) The tests proved that the trial wheel with the compound phase cut more freely, use less consumed power, and produce better finishes than conventional sintered diamond wheel with bronze-bond matrix.
Instead of conventional scanning methods, a new method for contour mapping a solid body is proposed. A surface tracing probe moves horizontally along a constant height on the solid, then the contour line at that height is obtained directly as a locus of the surface tracing probe. Therefore, the proposed method can be called a Direct Contour Mapping method (DCM). In comparison with the conventional scanning method, the DCM has following advantages : (1) local mapping is possible, (2) the storage of the height data of the solid can be minimized, and (3) the probe need not detect the absolute height. Owing to these advantages, the memory capacity and the time required for contour mapping by the DCM can be less than one-third of those by the scanning method. By using a micro-computer with Z 80 CPU, a direct contour mapping system is designed. A solid body to be measured is placed on an X-Y table, which is controlled by the micro-computer via GPIB bus. The probe of an electric micrometer attached to a computer controlled Z table traces the contour lines of the solid. Some examples of counter maps drawn by the system are shown.
This paper describes the proposal of a magnetic force sensor which enables a robot to follow a 3-D curved surface. The magnetic force sensor has a permanent magnet at one probe end and a three axial dynamometer at the other end. Since the magnet is attracted in the normal direction of the surface, which of course must be ferromagnetic, the inclination of the sensor axis from the normal can be detected by measuring the components of the off-axial force. Therefore, the robot can direct the sensor axis in the normal. Then the displacement from the surface can be obtained using the relation between the displacement and the attraction force which is now on-axial. In this way, the robot can follow the curved surface with the sensor axis normal to the surface, keeping the sensor end a constant distance away from the surface. In the present work, characteristics of the sensor are investigated, and measurement of a surface contour is tried using an articulated robot installed with the magnetic force sensor.
The point of view of digital filter was applied to the problem of surface undulation measurement. As a result the error was discussed in spatial frequency domain, of tangent method, inclination method and their modification. In spatial frequency domain, error of measurement is estimated clearly. Inclination method correspond to the measurement that leg width is equal to sampling interval and integrated by midpoint formula. The method except condition one by one from inclination method means to 2 points method. 2 points method opened up a prospect that avoid the probrem of aliasing. Knowledge of transfer function makes elevation of accuracy possible by inverse filtering. In case that data is uncontinuous at both ends, insert data made by spline function restrain of Gibbs phenomenon and makes high accuracy.
This paper deals with large deflections of thin, straight and tapered cantilever beams set up almost vertically. The results are as follows : (1) When a cantilever beam is placed almost vertically, the deformation of the beam is very small for the load lower than a certain critical value, while it becomes very large suddenly for the load exceeding the said critical value. (2) When the inclination of the beam exceeds the vertical angle, two deformation forms can exist for the load greater than the above-mentioned critical value. As the inclination increases gradually from a small value, the deformation form changes continuously, and when the inclination reaches a certain value exceeding an angle of 90°, it jumps from the 1st form to the 2nd. (3) When the load is applied slightly off-centered and more over the inclination passes through an angle of 90°, the beam turns to the direction of the eccentric load and the deformation form transfers from the 1st form to the 2nd. (4) In the actual fishing, it is experimentally known that if we incline the fishing rod by an angle of more than 90°, we often lose the control of the rod because the rod tends to turn clockwise or counterclockwise. This fact can be explained by this dynamical analysis.
A conceptual constitution of general wear simulation is discussed fundamentally, and a wear simulation system for trapezoid leadscrew is constituted. Conditions which rule wear phenomenon are classified into inside conditions, mechanical conditions and outside conditions. And a wear process is considered as a system which transforms such conditions into evaluation variables of wear. The wear simulation with geometrically simplified models reproduces the local phenomenon of the practical system, and it was assumed that a wear phenomenon is determined by local conditions. Considering the similarity rule on a wear process, it becomes obvious that the contact area ratio and all the local conditions of the simulation model needs to remain identical with those of the practical system. On the basis of these considerations, a wear simulation system for trapezoid leadscrew using cylindrical model is constituted conceptually, and procedure of simulation is given.
Three kinds of mechanical models for rubber vibration isolators are analyzed, and their transmissibilities are compared with experimental results. Two-elements model which is known as Voigt model is widly used, but its calculated value of resonance frequency does not agree with the measured one. Three-elements model is composed of two spring elements and one damping element. The characteristics of resonance can be made to agree with experimental results by using proper values of three constants. However, when the mass of the load is changed, the calculated characteristics of resonance do not agree with the measured ones. Four-elements model is composed of two spring elements and two damping elements. In this case by using proper values of the four constants, the calculated transmissibility characteristics can be made to agree with the measured ones for the wide range of load mass. It is concluded that the four-elements model is most suitable for the rubber vibration isolator.
In present study, the authors propose the new method to find the pressure distribution acting on contact surface and develop the measuring device. The principle of pressure measuring is based on the scattered surface reflection of light from the surface after contact. With this device, a pressure distribution is measured automatically and displayed in three dimension. It is shown pressure distributions in various contact cases measured actually using the developed device.
The study presents the method of automatically measuring dimensions of unknown workpieces placed at an arbitrary position by use of a 3-dimensional coordinate measuring machine (3 D-CMM). To recognize the workpiece shape, the image processing technique is utilized. The recognition of workpiece shapes leads to the generation of measuring paths of 3 D-CMM. Though the workpiece shapes are limited to the combination with rectangular prisms, the method is found to have the potential of automatically measuring workpiece dimensions.
This paper presents the method for predicting the maximum axial force Fm of the bolt induced by tensile impact energy E0 under the various conditions of the bolted joint with plastics bolt. First, the influence of clamped member and initial tightening force on the Fm is investigated experimentally and prediction method that corrects the result of elementary test to a value corresponding to any given fastening condition is proposed. Second, the influence of different screw size and clamped length on the Fm is considered for the purpose of generalizing this method. As the result, obtained conclusions are summarized as follows : (1) Influence of the spring constant of practical clamped member on the Fm is negligible as compared with that of initial tightening force. (2) The proposed prediction method is applicable to the bolted joint with different screw size and clamped length by generalizing the Fm by stress and the E0 by impact energy per unit volume of the bolt.
According to the proposal for ISO 1328 on the basis of DIN, the evaluation of tooth profile error requires the “mean profile” which is a modified version of the design profile. However, the determination of a mean profile has some troublesome problems. A simple and correct method to determine it by the method of least squares is presented. The determination of a mean profile can be carried out through the following three steps: (1) The least-squares fitted line of the design profile is obtained by the method of least squares. (2) The least-squares fitted line of the actual (measured) profile is also obtained. (3) The mean profile is provided as the sum of the following two components. The one is the slope component given as the least-squares fitted line of the actual profile. The other is the form component which is obtained by subtructing the least-squares fitted line of the design profile from the design profile itself. A prototype system, which consists of a personal computer and a digitizer, developed due to the above mentioned principle.
In order to realize straight motion at the highest level of accuracy, a system is introduced in which five degrees of freedom of a moving table; vertical and horizontal positions, and pitch, roll and yaw angles, are controlled precisely. The table is supported by active air bearings composed of PZT actuators and air pads. A PID-PD controller is designed achieving the table motion resolution of 0.02 μm and 0.02 arc sec. Errors of two degrees of freedom; horizontal position and yaw angle, were measured by a straight edge master, profile error of which was eliminated using the reversal method. Absolute accuracies of controlled straight motion were 0.14 μm in horizontal position and 0.14 arc sec in yaw angle over 600 mm travel.
In this paper, changes of cutting mechanism in the plasma hot machining of 18% Mn steel are analyzed by means of a two-dimensional cutting simulator incorporated with the flow stress characteristics measured. Firstly, the outline of the simulator is presented, and then the material properties and simulation models used in the analysis are described. Experimental observations, such as the decreases in cutting forces and shear angle, the increases in chip curl radius and chip contact length in the hot machining are successfully simulated. These changing phenomena can closely be related to the changes of strain rate, strain, temperature and flow stress distributions within the primary deformation zone. Especially, the drop of flow stress at the free-surface sides of workpiece and chip by the heating is a main cause of the changing mechanism. Under an ideal heating condition, in which the heat penetrates up to the undeformed chip thickness, the temperatures on the rake face and finished surface are almost independent of the heating. Relating this fact to a wear equation proposed elesewhere, less cratering in the hot machining can analytically be predicted, which is due to the decrease in normal stress on the rake face.
This paper presents some experimental results on mechano-chemical polishing of ferrite, and suggestion on the mechanism of mechano-chemical polishing. Such elastic polishing plates as synthetic leather is popularly used with an alkaline polishing fluid in the mechano-chemical polishing of semiconductive materials. In this paper, however, polishing plates made by either tin or lead are used with chemically active polishing fluid to ferrite. Electro-chemical action is added to mechanical polishing for the sake of composing the local cell between a polishing plate and a workpiece. Electro-chemical action not only increases stock removal rate twice or more, but also reduces the surface damaged layer of workpieces a quarter or less in comparison with those in mechanical polishing process.
The geometrical meaning of the optical characteristic roughness SN is discussed for the purpose of the use as the scale parameter to estimate the quality of the belt sanded surface. The experiments were carried out under the various grinding conditions by the platen type belt grinder. The ground surface is constructed by the long crested asperities which will be important estimation for quality. The surface roughness parameters such as density of summits, the mean curvature and the mean slope of asperities, the number of peaks and the zero-crossings were calculated by the profile and surface spectral moments. The functional relationship between an SN and the geometrical roughness parameters were derived statistically by the regression analysis. It was found that an SN has strongly correlated with the surface invariants of the spectral moment which indicate the curvature and the slope of the asperity at the same time.
In order to reduce the tool wear in cutting GFRP, turning of both pipe and disk of GFRP and drilling of a GFRP plate were carried out under the condition of elevated temperature below 423 K. The pipe was the unidirectional glass fiber reinforced polyester resin, and the disk as well as the plate was the glass cloth reinforced epoxy resin. Both the mechanism of reducing the tool wear and the influence of the cutting temperature on it are considered from the experimental results. The reducing effects of tool wear range from the low cutting speed inducing the relative small tool wear to the high cutting speed inducing the large tool wear directly proportional to the cutting speed. This is based on the decrease of the shear strength and the hardness of the glass fiber, and the decrease of the viscosity of matrix. However, as the bonding between the glass fiber and the matrix also becomes weak at elevated temperature, the condition of the cut surface is not good in many cases. The cemented carbide P 10 showed the remarkable increase of tool wear even at 373 K in heating the tool itself. Consequently, in cutting GFRP with the carbide tool, it is considered that the cutting heat has the two effects, that is, the one is the reducing effect of tool wear due to the variation of the mechanical properties of workpiece, the other is the promoting effect of the tool wear due to the decrease of hardness of the bonding phase of carbide tool and these two effects cancel out. In case of coating tool, the flank wear width decreases with increase of cutting speed. This differs from the case of the cemented carbide tool, the cermet tool and HSS tool.
The measuring accuracy of grinding wheel surface topography is investigated by Monte Carlo simulation for the following five indirect measuring methods from the viewpoint that the elastic and plastic behaviors in the grain-work interface may have an influence on the measured density of grain cutting edges : razor-blade method, thermo-couple method, scratch method, cluster overcut fly grinding (COFG) method, and conventional grinding method. From the simulation, it is clarified that the number of grain cutting edges obtained from the razor-blade method, in which total effective cutting edges are the targets to be measured, and that obtained from the thermo-couple method, in which effective cutting edges on the reference section are the targets, are more influenced by the elastic and plastic behaviors than those obtained from COFG method, scratch method, and conventional grinding method, in which effective cutting edges for roughness generation are the targets. In addition to the simulation, the following is verified by some experiments : the number of grain cutting edges seized by the thermo-couple method is larger than that seized by conventional grinding method.
This paper describes a new method for inserting operations of the parts with the clearance of several micrometers. In this work, it was found that inserting motion became easy by the application of ultrasonic vibration to one of the peg or the hole. This effect is caused by the reduction of the apparent friction on the surface of rod (in this case, that is the peg), the generation of the radial vibration owing to the Poisson's effect and the force directed to the node position of rod, which is generated by the combined vibration of the longitudinal vibration and the radial vibration mentioned above. Besides, as for this inserting mechanism, periodical impact was applied to the rod which has any kind of length with natural frequency of its own. It was found that the rod vibrates at its own natural frequency or at the frequency of several times of the natural frequency. Then, a prot-type automatic inserting system with a scara type robot was developed. The superposition effect of vibration onto the parts to be assembled was confirmed.