[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

Computer aided automation was developed for designing and manufacturing frame members of the wooden Japanese house built in the traditional style of construction. The total system includes a computer system which process the design and manufacturing data, a data station that takes care of the data transfer and an automatic cutting machine, so-called "Precut Machine" which performs the machining of the frame members. For processing the design and manufacturing data, a data structure was proposed for classification and storing the input data which is prepared in an interactive graphic manner. Furthermore, based on the survey and discussion on the various existing patterns of connecting frame members, computation algorithms and data structures were developed so that a connecting pattern and its corresponding machining program can be automatically generated for all the connections to be worked for building a given house structure. System software was implemented in a cost effective personal computer system. By linking the personal computer with automatic cutting machines via the selfdeveloped microprocessor-based data station, an evaluation test was performed. Results have shown a remarkable increase in productivity of designing and manufacturing members of wooden Japanese houses.

View full abstract

In view of getting a microcomputer with high efficiency at a reasonable price it is intended to convert an existing NC lathe into the Behind-the-Tape Reader DNC system by use of a microcomputer. This will make the most of the existing NC lathe and bear comparison with CNC lathe equipped with sophisticated functions. The interface is designed to link the tape reader of NC controller to the microcomputer. The automatic programming software is also developed to automatically create NC instructions by giving data related to the configuration of a workpiece, tools, and cutting conditions, interactively with the microcomputer. As a result, it becomes apparent that the BTR-DNC system, by use of the inexpensive microcomputer, works without any trouble and makes the conventional NC machining remarkably convenient and comfortable. The ability of reconstructing existing NC controllers to DNC systems leads to the effective use of obsolecent hard-wired NC controllers.

View full abstract

In the course of developing part specifications from assembly drawing, several operations such as the verification of assembly drawing, the examination of the possibility of assembly, the determination of sequence of assembly/disassembly, and the distribution of tolerance are to be executed. To aid this process, which is carried out by computer, the assembly structure of parts should be formulated as a data model. This paper describes the following two subjects concerning the above :(i) a data model of assembly structure data by means of connective relations between the parts of functional machine units with their properties;(ii) an algorithm on the sequence of assembly/disassembly of the functional machine units, according to the priority of the connective relations, assuming that the sequence of assembly is the reverse to that of disassembly. The data model and the algorithm of the sequence of assembly are proved to be practically effective through case studies of the functional units of machine tools.

View full abstract

This paper deals with the high-speed CP motion of a robot. When actuating robots at high speed, non-linear forces, such as inertia force, Coriolis force, centrifugal force, must be taken into consideration. And the driving torques to compensate such non-linear forces are desired to be smooth and small. As the result of the simulation of path planning by some typical spline functions, it has been proved that the best path planning method is to produce the trajectory of a joint whose torque or angle is exceptionally large, by 3rd order polynomial, and to produce the trajectories of other joints in relation to the studied joint. Next, to test the actual effect of this method, an experiment of dynamic control was carried out. The experiment proved this path planning method to be suitable for the high-speed CP motion.

View full abstract

In this paper, the load distribution formula on a ball spline is derived in the case where a radial, a moment and a torque load are simultaneously applied to the ball spline. The fluctuation states of the load distribution and of the maximum distributed load are obtained in accordance with the ball movements in the ball spline and the basic static load ratings are made clear under various applications. The following is a brief summary of the calculation results : (1) The fluctuation state of maximum distributed loads was very extreme and it showed a peak value. Therefore, it is found that the basic static load rating for ball spline is to be decided on the load distribution state corresponding to this peak value of the maximum distributed load. (2) In the case of moment loading the load distribution state in the ball spline became uneven, while in the case of torque loading it became even. As the basic static load rating for the ball spline was dependent on the radial load, however, it was extremely decreased with the increase of the moment loading and/or the torque loading.

View full abstract

The methods of Rockwell and Rockwell superficial hardness tests have practical advantages of getting the hardness number with a small personal error and of making easily the testing procedure automatic, because the hardness number is directly shown on the indicator. The methods of Rockwell quarter load and Rockwell quarter load superficial hardness tests are proposed in order to apply these advantages to test a small or thin specimen. As compared with the usual methods, the indenters of half size and a quarter amount of the loads are applied, and then the coefficients of the depth of indentation in the calculation formulas for hardness number are made twice. Even if the test load is changed, the same Rockwell hardness number is obtained when the indentation is geometrically similar under the condition where the ratio of the load to square amount of the size of the indenter is constant. To confirm this theory of similarity in hardness by experiment, the Rockwell and Rockwell superficial hardness testing machines and respective quarter load machines are produced. The errors of these machines' elements are integrated in HRB or HR 30 T as follows : their algebraic sums are within -0.5-0.3, and the roots of the sum of their squares are lower than 0.3.

View full abstract

A deflection of an internal gear with thin rim is composed of both the deflection of tooth and that of rim. Therefore an equation of deflection of a tooth of the internal gear is obtained by replacing the tooth by a cantilever composed of two trapezoid and an equation of deflection of a rim of the internal gear is obtained by replacing the internal gear by an equivalent ring. Then using the equations mentioned above and an equation of deflection of an external gear, which is widely used, the stiffness of a pair of teeth between the internal gear and a planet gear is illustrated against the number of teeth of the internal gear and that of the planet gear, and also the stiffness of the rim is illustrated against the number of planet gears, the number of teeth of the internal gear and the rim thickness. Further a load sharing rate of the internal gear with thin rim is discussed.

View full abstract

This paper deals with a method of estimating the force by dynamics on the feed mechanism of industrial sewing machine. As a result, the theoretical values obtained by solving the simultaneous linear motion equations with 24 unknowns agree with the experimental values. Especially, as for the force F_l produced in the feed forked connection lever, the followings are made clear. (1) The peak values of left and right component are produced at a nearly neutral position of the lever equivalent length l. (2) The fluctuation of F_l is accompanied with the higher harmonic of about 500 Hz by natural vibration of the transfer system in vertical direction to the moving plane during compression of the axial force F_f. In addition, as for the torque produced in the feed regulator, the followings are made clear. (3) During forward feed, this torque is the sum of the torques T_s due to spring and T_r due to inertia force. Therefore the set condition of the feed regulator is stable. (4) During backward feed, this torque is the difference. Therefore the feed reverse lever should be kept to be stable by the force exceeding the torque difference.

View full abstract

This paper describes development of a constitution that satisfies manifold needs for the designing system of profile shifted involute internal gears and also a simple, effective optimization technique (influence index number method) that can be applied to particularly complex programming such as required for designing of gears. The previous paper presented the system and technique for profile shifted involute crossed helical gears. This paper deals with internal gears. Internal gears are specifically subject to numbers of interference, which produces increased complication to the effect of designing variables. Especially adequate consideration is, therefore, required. The system and technique developed were fully tested by extensive actual design computation and trial manufacture of an example of gears designed by the present method. The test confirmed that this designing system and simple technique were also very effective for designing of internal gears.

View full abstract

This paper aims to clarify comprehensively and quantitatively the effects of elastic deflection between wheel and workpiece, wheel wear and thermal deformation of workpiece on the profile and dimensional errors of workpiece which is produced in one pass creep feed grinding. The elastic deflection is determined from the experimentally obtained static stiffness between wheel and workpiece during grinding. Effects of grinding conditions on the wheel wear are examined experimentally, and it is shown theoretically and experimentally that the radial wheel wear depends almost solely on the metal removal. Workpiece temperature distribution is analyzed by using the finite element method and it is revealed to agree with the experimental result. From this the thermal deformation is calculated by the FEM. It is shown that the resultant errors coincide well with the actual profile and dimensional errors of workpieces. Effects of profiled wheel wear and the thermal deformation of workpiece being profile ground are also considered.

View full abstract

In order to compare with the tool wear in cutting GFRP, the glass beads reinforced polyester resin was turned. Wear characteristics of many tool materials were studied, and both the abraded surfaces of cutting tools and the glass beads in cut surfaces were observed with SEM. The tool wear is the abrasive one. The wear resistances of tool materials are similar to that in cutting GFRP. In some cases, the wear increases steeply with increase of cutting speed over a certain critical cutting speed. This phenomenon is similar to that in cutting GFRP. This occurs at the feed rate of 0.10 mm/rev in P 10 and Co-rich cemented carbide tools of WC-Co grade. In this case, the abraded surface of the tool is very rough because of random falling off of carbide grains and the glass beads in a cut surface rub the abraded surface lightly, so they scarecely fracture at low cutting speed. The abraded surface is comparatively smooth and the glass beads rub the abraded surface severly in the neighbourhood of the critical cutting speed. The abraded surface becomes rough again and the glass beads rub the abraded surface more severely, so many beads fracture over the critical cutting speed. This phenomenon do not occur at the feed rate of 0.05 and 0.20 mm/rev. This phenomenon is caused by the mechanism that the falling off of carbide grains and the fracture of glass beads in the cut surface promote each other over the critical cutting speed. This mechanism is quite different from that of the similar phenomenon in cutting GFRP.

View full abstract

In order to increase stability against machining chatter in boring operations, boring bars equipped with multiple degree-of-freedom dampers have been developed. By using the dampers, the modal flexibility at the natural mode having the greatest receptance is divided into smaller modal flexibilities at multiple degrees of natural modes. This theoretical analysis clarifies that the division of the modal flexibility into many natural modes results in 1) the decrease of the modal flexibility, 2) the increase of the damping ratio, and 3) shift of the vector locus of the receptance frequency response along the real axis toward the positive direction. These changes occur at the natural mode having the greatest modal flexibility, producing a great increase in stability against the machining chatter. Design procedures for the boring bars are described, and the effect of the dampers is demonstrated in actual cutting tests.

View full abstract

In order to examine the fracture characteristics as Markov process, a fracture test of WA single grain in a modified form of surface grinding is carried out. Using the instrument reported in the previous paper, created topography of grain cutting edge by the fracture is identified into the 10 representative models proposed and the transition probabilities between the models are obtained. The probability is found to be homogeneous and to satisfy the elementary Markov property. Concerning the fracture onset, it is found that exponential distribution always conforms with variation of the onset probability with number of the cutting, though the parameter value of the distribution depends upon the edge models. The fracture onset thus appears to also satisfy the requirements as Markov process. A formulation of analysis as Markov process to obtain variation of the edge shape distribution during grinding is proposed and some preliminary results of the analysis are discussed.

View full abstract

This paper deals with the tilt angle control, based on the tool oscillation and the effect of tool oscillation for curved surface polishing. The results are as follous. (1) The tilt angle of polished surface can be detected in polishing process from the relationship between tool oscillation and the displacement of a detecting shaft by using the oscillated polishing head. (2) The curved surface is kept horizontal or at a constant tilt angle in polishing process by controlling a tilting table using a detecting signal from the polishing head. (3) The surface quality polished by a orcillated polishing head is superior to that by a polishing head without oscillation, because tool contact area become wider when tool is oscillated.

View full abstract