In the optimization of injection molding process, the evaluation based on transformability has been suggested and it is fairly successful from the quality engineering view point. However, the technology may not be quite improved using this method. In a study of placental precise molding reported in this paper, test pieces of special dimensions were designed and two evaluating methods were compared: transformability of the shape and the use of proportional equation calculated from the weight in the air and the one in water. Reproducibility of the gain from the two evaluating methods were both satisfactory. But the optimum conditions were different depending on the case when the annealing process was or was not considered as the noise factor. It means that a passed (good) product inspected right after molding and the one inspected after time passage are not necessarily the same. In the study, the surface was measured in two ways and considered as a noise factor. This helped the application of transformability more effectively.

Improvement of a Rotation Rubber Belt for Copying Machines In the mid/high speed (40 to 100 copies per minute) copying machines manufactured by our company, belts are used to feed paper to the fixation process after the toner image exposed on the drum is transformed to the paper. When rotation is unstable, the belt moves to the front/inner direction, interfering other component parts and cause defectives. In this study, quality engineering approach was applied to stabilize the feed belt rotation.

The use of Computer Aided Engineering (CAE) for product development is more advantageous in reducing the development time cycle as well as reducing cost than the case of using hardware. However, a question is if the simulated results can be reproduced by hardware. In this study, the improvement of the rigidity of a structure without fabricating the hardware is reported. In the study, the rigidity of a structure was optimized using quality engineering and CGE. For the generic function, it was considered that the input (load) is proportional to the output (displacement) and zero-point proportional equation was used. The thicness of various dimensions was used and compounded as the noise factor. The standard SN ratio was calculated using the nominal amount of displacement as the signal. As a result, the collditions of improving the rigidity were determined and the reproducibility was confrmed. Moreover,the concentrated stress in the structure before study became dispersed after optimization. The study was completed in a short period of time at a low cost compared with the one using hardware.

SAW (Surface Acoustic Wave) devices were widely used as electric filters because of the small size advantage. In many cases, computer simulation is used to design SAW filters for the adjustment of frequency characteristics. In this report, quality engineering approach was applied to the computer simulation design. The first step was to stabilize frequency characteristic using the standard SN ratio, and the second step was to adjust frequency characteristic to the target. This study shows that quality engineering approach combined with computer simulation enables to improve the robustness of SAW filters. It also makes the tuning of frequency easier than the traditional design methods.

In our company, several in-house training courses for quality engineering have been conducted. At the beginning of introducing quality engineering, it is important to learn the concepts of dynamic characteristic function and parameter design. For this purpose, the training materials were developed including a series of parameter design steps and the confirmatory experiment using the spring driyen mini-cars purchased from the market. The training materials were designed to achieve the following objectives:

(1) To break the barrier of accepting new concepts.

(2) To learn and practice dynamic characteristics.

(3) To learn the steps of parameter design.

(4) To experience the reproducibility of experimentation.

Seven inter-changeable components with different shapes and materials in a car were used as major control factors. Spring winding distance and the running distance were used as the signal and output characteristic respectively. From the results of application, it was found that the materials were useful to help understand various concepts such as the system. function, control factor,quality engineering, dynamic characteristics, reproducibility, variability, orthogonal arrays and practicing of parameter design. It was also found that the attendees were satisfied. The training materials will be further improved in the near future.

Optimization of the Inorganic Film Forming Condition Using the Liquid-Phase Deposition Method In the liquid-phase deposition process, inorganic film is formed on the glass flake. Parameter design was conducted to optimize the film forming condition using a beaker scale experiment before the construction of the production equipment. In the experiment, the time of film formation was considered the signal factor, the thickness of formed oxide film as the output, and flake concentration as the noise factor. Since the measurement of film thickness on the minute flakes was very difficult, the film tone was converted into thicness by providing a conversion line. After the parameter design, a gain of 9.12 db was confirmed. From the result, it was possible to design and construct the production equipment with the minimum cost. Since the large scale experiment was minimized, research cycle time was greatly reduced.