This paper proposes design decision making methods of machine structures based on design sensitivity analyses for minimizing the influence of variations in the manufacturing processes on the structural characteristics (product performance). Design decisions are made so that smaller design sensitivity coefficients of the characteristics are given for areas that have been estimated to have a great potential for variation in the design variables. First, fundamental design principles are established for minimizing variations of the characteristics. Then, the practical procedures are presented for decreasing the variations of the product performance. The proposed method in which the minimization of variations of the product performance is the main objective corresponds to a maximization of the product performance and also brings about a decrease of the manufacturing cost. Finally, the effectiveness of the design decision making methods are demonstrated for designing a machine tool structure.
An experienced process engineer immediately judges in incredibly short time regardless of the complexity of contents described in a given part drawing in case of actual operation. According to our investigation performed previously, this is because the engineer is carrying out the process planning of flair-based type, in which a diagram image of a part drawing is used preferably with employing the top down-like procedure. Then our former paper suggested a general concept of CAPP system of flair-based type. In this paper, a method of knowledge process is proposed by the structural matching technique based on the imitation of model patterns of process plan-examples, and applied this method applies to the simple case of the process planning problem.
In centerless grinding, unbalanced parts are hard to get precise accuracy, because the centrifugal force caused by rotating the unbalanced mass tends to move the part up and down, and/or fall down the part from work support blade. The unbalanced parts are classified into 3 types by the position of unbalanced mass. The first report presents the criteria of this movement of each type of the unbalanced part by analyzing the balance of the forces working on it. This paper deals the experiment of centerless grinding of those 3 types unbalanced parts. The experimental results indicate that the accuracy was not affected substancially by the unbalanced mass under the stable grinding conditions, and this is coinsident with analytical results.
The geometrical surface generation is simulated for the grinding process with the grinding speed ranging from 25 to 1000 m/s. The effects of work speed, depth of cut, spark-out cycles, and cutting edge density etc. on the surface profile and the surface roughness are investigated. The three-dimensional surface profile generated by the cutting edges is described geometrically for each cross section of the workpiece which is perpendicular to the direction of grinding, considering the distribution of the cutting edges in a grinding wheel. The simulated results are displayed on the CRT by the computer graphics, and the surface roughness is also estimated theoretically. As the results, the following conclusions are obtained: (1) The high-speed grinding reduces the surface roughness, comparing with conventional grinding. (2) The work speed and the cutting edge density have little influence on the surface roughness in the high-speed grinding. (3) Relatively good surface roughness can be obtained by the high-speed grinding with a few spark-out cycles, regardless of pre-surface roughness.
A new efficient magnetic abrasive finishing process is created in which the new type magnetic abrasives are used mixing small sized diamond magnetic abrasives with such large sized ferromagnetic particles as iron particles. From the experiments, it was clarified that the machining efficiency was improved remakably based on the increasing effects of both the finishing pressure and the magnetic brush rigidity. It was also pointed out that the machining efficiency was influenced severely by both the iron particle diameter and the mixing weight percentage of iron particles, and there was suitable value for the mixing weight percentage.
A fuzzy rule-based expert system currently is being developed whose goal is the automation of decision about the five factors of grinding wheel. The decision of those factors is dependent on the dictates of experience, because the constraints and the objective function are little known. Therefore, a heuristic approach to the decision problem has been employed. This paper illustrates the designer's knowledge on the selection from the great variety of the abrasive types, the grain sizes, the grades, the structures and the bond types, and describes the fuzzy logic framed in a way of representing his knowledge as the rule for the use of machine reasoning. Furthermore, the paper describes the fuzzy decisional algorithm used in the inference.
Some machine parts, such as cutting tools and shearing blades, have importance in their corner shapes. The corner dullness which deteriorates the corner shapes is often generated at the ends of ground machine parts. This paper deals, both theoretically and experimentally, the mechanism of generating the dullness under one-pass-plunge grinding and following results are obtained. (1) Entrance dullness, being generated at the wheel entrance part, and exit dullness, being generated at the wheel exit part, are different in shape. The height of exit dullness is about three times as much as the entrance. (2) Shapes of the dullness, theoret-ically obtained under consideration of the wheel-workpiece relative displacement and utilizing the one-freedom vibration model, show good agreement with measured results. (3) Height of the exit dullness increases with wheel depth of cut, table speed, and grinding stiffness under usual grinding condition. (4) Height of exit dullness decreases with equivalent mass, dumping coefficient, and spring constant of the grinding machine.
Tool entry process in three dimensional cutting is simulated by finite element method. Since the chip forming process is one of large deformation problems, geometrical nonlinearity due to the change of shape is introduced into finite element formulation. Constant strain tetrahedra element is used and over-constraint of incompressibility on the deformation of the finite elements in plastic region is relaxed using Nagategaal-Rice functional to make the elements as deformable as real continuum. The chip formation from tool entry to steady state is realized in the analysis and stress and strain distributions in chip and machined surfaces are also obtained. Lower shear angle at side surface of the work, side flow and chip flow direction predicted are in good agreement with experimental results of cutting of plasticine.
The slack-of-belt grinding method is used in the finishing process of high-grade furnishings and parts used for special functions. However, the abrasive belt flexibility causes difficulty in recognizing its characteristics and in automating this finishing process. In this study, for automation of the slack-of-belt type grinding process using fuzzy theory, fundamental research was carried out. A fuzzy rule was derived by investigating the characteristics of the slack-ofbelt and by arranging the experimental results. This control system keeps the amount of stock removal at the required values. This system infers the necessary grinding time from the abrasive belt conditions by employing the fuzzy rule. The belt conditions are judged from the tangential grinding force which is detected when the workpiece is pressed against the abrasive belt. As a result of the experiments, it was found that there was a great possibility of applying the fuzzy theory to the automation of the grinding process using an abrasive belt.
A computer simulation system is developed to estimate the heat affected zone (HAZ) of hypoeutectoid steel in surface grinding process. HAZ is generated by the liquid-solid phase transformation and solid state phase transformation caused by the grinding temperature. In the present system, the heat flow and phase transformation during grinding process is analyzed using a finitedifference method. The changes with the time of temperature distribution, the carbon redistribution in austenite, melting and the subsequent solidification to dendrite, and the quenching to martensite are simulated on the EWS. The analyzed results of the distributions of temperature, microstructure and carbon concentration are shown in the pseudocolor images on the display. The microstructure and the hardness in the heat-affected zone after the grinding process can be predicted. The present system can be applied to determine theoretically the best combination of the process variables for the grinding in advance.
It is important to know the temperature distribution in a workpiece to understand the curvature of thin plate in surface grinding. In this paper, the relationship between curvature and calculated temperature distribution by FEM was clarified. As a result, it is evident that the steeper temperature gradient the larger concave curvature is induced. This can be understood as an effect of thermal stress. In the case of gentle temperature gradient, the effect of thermal stress is not clear, and consequently burnishing by cutting edge becomes effective. Summarizing these results, the curvature is well explained with temperature gradient and burnishing effect. The transformation of 304 stainless steel is confirmed by means of X-ray diffraction test, and the influence of transformation on curvature is studied. Key words: surface grinding;superabrasive wheels;thin plate;curvature, temperature distribution;force ratio;transformation.
reep-feed grinding is a form-copying grinding process capable higher rates of material stock removal than most other conventional grinding processes. In industry, creep-feed grinding is used for form-copying grinding of material which is very hard or very tough. The most widespread application of the process is the form-copying grinding of Inconel turbine blades of gas turbines. This paper deals with the grinding wheel speed and grinding ratio aspects of creep-feed grinding of Inconel 738. These aspects were studied under conventional grinding and creep-feed grinding conditions. The following results were obtained. (1) In grinding conditions, the grinding ratio has an especially strong correlation with the grinding speed. The grinding ratio at a low wheel speed of 3 m/s is approximately 10 times larger than that of the conventional wheel speed of 33 m/s. (2) It is made clear that dulling and breaking appear on a grinding wheel at a conventional wheel speed, and self-sharpening appears at a low wheel speed, according to the estimation with the measured mean grain profiles. (3) The wheel wear of the edge part during creep-feed grinding is 4 times less than that of the conventional grinding, with low wheel speed operation.
Etching mechanism and characteristics of CW-CO2-laser induced thermo-chemical etching of ceramics (ZrO2) in halogen-containing gas (CF4) were investigated. It was presumed that the etching velocity depends on activation energy of gas molecule dessociation and laser irradiated surface temperature. From the computer simulation of the etching process and the experiment, following results were obtained. The etching process can be induced on the material surface under the melting point. Depth of the thermal affected layer of the etched surface can be controlled by the etching parameter, activation energy, laser power and laser irradiation time, so laser machining of ceramics without thermal affected layer is possible.
FIAC-III System (Fuzzy Intelligent & Automatic Control III System) is developed in this report in order to carry out the optimum grinding by neuro & fuzzy in-process control of infeed rate and speed ratio. In this system final surface roughness and grinding efficiency are optimized by means of in-process prediction of surface roughness by the neural network and rapid retraction of wheel head just before spark-out. Main conclusions obtained are as follows : (1) Neural network model for prediction of surface roughness is built up based on the back propagation method. As a result, it is possible to predict surface roughness in grinding process. (2) In FIAC-III grinding, the final surface roughness is satisfied with the required value independent of the wheel surface condition, and the grinding efficiency becomes much better than that of the traditional one plunge speed grinding.
It is important for precision profile measurements on on-machine conditions to construct some accurate datum for scanning. Any preceding method using software datum is not possible to measure step-wise profiles accurately. In the inclination method, it is necessary to hold some distance between two displacement sensors and in most cases the distance is too wide to represent a sharp edged stepwise profile. The generalized 3-probe method and other methods using software datum can represent more higher frequency components than the inclination method. However the shape of step can not yet be expressed because of Gibbs's phenomenon in the Fourier transform which is used for deconvolution. In the present paper it is proposed to combine the inclination method and 2- or 3-probe method. The new method determines the relative height between the discrete data groups of the inclination method by using generalized 2- or 3-probe method. And the new combining method can maintain the advantage of the inclination method that the discrete data present rigorously the original relative height of sampled points without any errors of data processing.
Image processing using neural networks is applied to the inspection of electronic connector plug pins. Some defects among a large number of pins inserted into a board are detected by the proposed system instead of visual observation. Pin images taken by ITV camera are first processed to digital binary signals, then input to the first layer of three-layered neural network. The training of the network is accomplished with the back-propagation algorithm. Each weight interconnecting units of adjoining layers is modified by learning many types of pin conditions. Fundamental features of the proposed system were made clear, and the network could well detect such defects as bended, mis-positioned and buckled pins.
This paper describes studies on the alignment error budgets caused by a TTL (through the lens) alignment optical system (SMART : separated mark TTL alignment) for an excimer laser aligner. Alignment error factors, such as the optical phase fluctuation of the alignment beam, the inclination variation of the alignment incident beam, the tilt of the wafer and the reticle, and the defocus of the reticle, have been analyzed and estimated. It was demonstrated that the total error of the alignment optical system was required to be better than 0.02 μm in order to satisfy the positioning strategy of the excimer laser aligner with a 0.05 μm (3σ) overlay accuracy. Several compensation methods have been developed and adopted to SMART. As a result, it was obtained experimentally that the total error of the alignment optical system became better than 0.02 μm. It has met the development goal. The new SMART optical system which was adopted to the KrF excimer laser aligner showed an excellent performance with a better than 0.05 μm (3σ) overlay accuracy.
For determining the presence of adsorbed material on surfaces of processed materials, an interferometer is developed using 325-nm He-Cd and 633-nm He-Ne laser sources. A measurement resolution of 0.5 nm is realized by scanning the interference fringes with a translator and taking the correlation between the probe fringes and the reference fringes. From experiments with many types of materials, it is found that the interferometer can detect adsorbed material on the surfaces of materials such as metals.
This paper presents a mathematical model of transient response of rotating velocity in a stepping motor driven with a constant frequency. The model is required because it can indicates analytically the pattern of transient response in the conventional mechanical vibration theory. A derivative equation with respect to small deviation of the angular position from the steady state results in Hill's or Mathieu's equation well-known as parametric excitation. From the solutions of the present model, the damped frequency and the amplitude damped ratio for the transient response of rotating velocity are analytically investigated. Furthermore, the calculated results and experimental results by various loads and switching frequencies are compared.
This paper describes the new feed control method-the modified inverse transfer function control method-of the numerically controlled machine tool to guarantee the geometrical accuracy of three-dimensional surface at very high feed speed.The method makes it possible to eliminate theoretically the feed servo error which is generated in feedback control and feedforward servo control. The control devices developed by the method consist of five 32 bits microcomputers which read numerical control command in advance, make interpolation for smoothing by hzier curve, calculate constantly acceleration and feed rate, control the feed rate within a given acceleration and generate the numerical control final command in real time by multiplying inverse transfer functon. As a result, it was confirmed for three axis numerically controlled machining center tokeep the geometrical accuracy of free surface workpiece weighing 98kN within ±0.02mm at feed rate 10m/min.