This paper reports how quality engineering simulation results were evaluated to develop a stabilized shutter mechanism. Since the shutter mechanism in question has no feedback control system, the product including the mechanism must operate at a minimum cost. To develop such a mechanism, stabilization factors were sought by evaluating results of quality engineering simulations to develop a simple shutter mechanism less susceptible to environmental changes. The stabilization factors were successfully tracked down by comparing the S/N ratio determined on the assumption of standard input signals (standard S/N ratio) with the S/N ratio found by conventional methods of analysis.

For the evaluation of various types of liver disease, three types of unit spaces differing in the numbers of subjects and items were constructed and Mahalanobis distances for patients in different disease levels were calculated. In unit space I, 30 subjects were evaluated as to 16 health examination items, including 9 ordinary items and 7 additional items; unit space II had 593 subjects and only the 9 ordinary items; unit space III had the same 9 items and 30 subjects. Unit spaces I and II both provided adequate discrimination between normal and pathological subjects based on Mahalanobis distances. The dynamic characteristic S/N ratio was determined by using disease level as a signal factor,and unit spaces II and III were compared by using the orthogonal array item selection method. The gain of the S/N ratio was found to be higher in unit space II than in unit space III, and for 9 items, the unit space with more subjects was more accurate. Another comparison of S/N ratios was made using score values and the logarithm of the Mahalanobis distance as signal factors. Use of score values increased the S/N ratio. These results demonstrated the reliability of unit space II, and suggested that the use of a unit space the reliability of which has been verified by the S/N ratio enables more accurate diagnosis.

Transferability of resin injection molding was evaluated by simulation. Ratios of mold dimensions to product dimensions were used as evaluation characteris tics. Gain reproducibility was low when the vertical, horizontal, and diagonal dimensions of the main parts in the design drawings were evaluated, but increased when more dimensions were evaluated by taking combinations of the above. The stability of a plane determined by three different points on a mold was evaluated by the angle formed by the normal vector to the plane, but the gain was not reproducible, and in all of the evaluations above, characteristic values did not vary uniformly with respect to noise: reversals of increase and decrease in the dimensions occurred. This was thought to be the cause of the low gain reproducibility. It would be wise to assign noise factors to an outer array and evaluate all combinations to improve the gain reproducibility.

Steel parts for automobiles are mainly made by forging. A shot blast process is also carried out to remove oxidized layers generated in the forging process. The shot blast process is in principle the same as the shot peening process that is used to increase fatigue strength of materials. The shot blast process parameters were therefore redesigned so that the shot peening effect could be obtained by using shot blast equipment. This improvement had an economic impact of 280 million yen per year.

The two most important fuctions of an organic photoconductor (OPC) are charge transport and photoconductivity decay. The following factors were used to evaluate charge and photoconductivity decay curves so that OPCs exhibiting the best performance in these functions could be manufactured under optimized conditions. Signal factors: charge supplied to the surface of the OPC and exposure time. Noise factors: electrostatic fatigue by repeated plain paper copy (PPC) operation and changes in OPC characteristics with time. Control factors:drying conditions for each layer of the OPC in the manufacturing process. The optimal conditions were defined as the conditions under which the S/N ratio and electrical sensitivity were equal to those obtained under conventional con ditions and the highest production speed could be obtained. The above opti mization method reduced spending on new production equipment, increased productivity, and lowered manufacturing cost.

Optimal conditions for processing special-purpose glass with high-speed machining centers were sought by using standard glass test pieces. Efforts were made to evaluate processing conditions based on the relationships between power consumption and processing time and between power consumption and mass of material removed, but many test pieces broke before data could be obtained. To solve this problem, processing progress, power consumption, and amount of material cut were ranked; the optimal processing conditions could then be identified.

To shorten schedules for the production of ever smaller parts made of increasingly refractory materials with increasingly complex shapes, the development of efficient processing techniques has become an important issue in the metal pro cessing field. The installation of equipment that can feed and process metal at high speed is considered an effective way to achieve high precision while short ening machining times and simplifying the machining process. Practical machining conditions often require lengthy optimization for each control factor,including tool selection, feed rate, and cutting speed. The present study undertook the optimization of microdrilling parameters by using quality engineering techniques to evaluate the stability of the machining process. Tests were conducted by carrying out a semi-dry drilling process with high-speed machining equipment, using an ultra-rigid solid drill with a diameter of 0.2mm. Provisional results were obtained for each control factor. The processing time per measured step quantity was found to differ by an order of magnitude from the values calculated from the nominal rotational speed and feed rate.

Users of copiers expect to obtain output clearer than the original image. Recently deyeloped digital copying technology has become highly flexible in adjustment of output images, providing features to satisfy users' demands down to the last detail. However, it is difficult for ordinary users to make the necessary fine adjustments by themselves. This study was carried out to develop an output image adjustment technology by which copiers can automatically classify original images into groups and make optimized adjustments that take best advantage of the copiers' advanced adjustment capabilities without troubling the user. Application of the MT system to the classification of original images led to a promising outlook for a high-precision classification system. Application of this technology will make it possible to implement a new generation of copiers that can exhibit the best features of digital copying technology while remaining easy to use.