To improve the quality of spot welding, following optimization of spot welding conditions in 2003 and spot electrode dresser cutting conditions in 2004, a spot electrode dresser functionality evaluation was carried out in 2005. As welding has become more difficult due changes in body steel plate material and changes in the thickness ratio of inner and outer panels, however, there are spots that cannot be welded well with the optimal conditions derived from those past studies. In the present study, past activities were reviewed and generic functions,control factors, and noise factors were reconsidered. It was decided to continue using energy evaluation, as in the previous studies. The gain was estimated to be 9.32 db, and was 7.11 db in a confirmatory experiment. Similar effects were obtained when the results were introduced onto actual production lines.

Molded parts containing photoluminescent materials are expected to find numerous uses because they have high installation flexibility and long service life. However, as photoluminescent materials are very expellsive, it is necessary to develop molded parts that can maintain light emission for extended times with as small an amount of photoluminescent material as possible. An improved set of molding conditions was obtained from an earlier optimization experiment with control factors allocated oll an orthogonal L₁₈ array. As the reproducibility of the gain was low, however, it was necessary to reexamine the functional definition and evaluation method. The experimental plan was therefore reconsidered with the aim of finding still better conditions. In the previous experiment the light emission function was evaluated. This time, the light storage function was also evaluated. As a result of the new experiment on the light emission function, reproducibility of the gain was obtained and better molding conditions were found. However, reproducibility was not obtained in the experiment on the light storage function, so it will be necessary to reexamine the evaluation method.

Generic control functions were studied in relation to PID feedback control. It was noted that the block diagram of a PID control system is basically equivalent to the damped oscillation model of a mechanical system with a spring, mass, and damper, and that energy conversion in PID control is a conversion of potential energy, the value of which is given as a change of the target quantity to thermal energy. However, thermal energy is difficult to measure. A residual energy was therefore defined as a sum of potential energy and kinetic energy. When the residual energy becomes zero, it can be considered that all the energy put into the system has been converted to thermal energy, so residual energy was evaluated by a smaller-the-better type of S/N ratio. Parameter design was carried out by this evaluation method. As a result, the S/N ratio improved substantially and good reproducibility of the gain was obtained. From these results it was concluded that when energy changes form but remains in the system,reproducibility of the gain can be secured by evaluating the entire residual energy in the system.

Research on electrical tools is aimed at making them smaller, lighter, and more powerful. In pursuit of these goals, the authors have worked on improving the performance of motors and optimizing their structural parts. A hedge trimmer,which is used to trim hedges and garden shrubs, is an electrical tool the performance and weight of which are largely determined by the cutting blades. In improving the blades, it has been customary to rely on the intuition and experience of leading tool manufacturers and designers. This time, the blade shape in a hedge trimmer was optimized with the goals of improved cutting performance and durability, and the further goal of a quantitative evaluation of cutting performance. Cutting performance was evaluated in terms of the standard S/N ratio by placing a static tension load on the blade and using the waveform of the energy applied to the blade. Blade evaluation, which had been a subjective art in the past, was successfully quantified.

In the post-assembly inspection of car audio systems to check for abnormal sounds, inspectors have been evaluating the sounds qualitatively. That is, then have been evaluating a quality characteristic by subjective human criteria. As these criteria vary from one inspector another, in some cases defective products are misjudged as normal and are shipped to customers. In this study, we tried evaluating operational sounds numerically by replacing the operational sounds themselves with a quantifiable acoustic pressure level and analyzing waveform patterns by the RT (Recognition Taguchi) method. By establishing fixed inspection criteria, we are hoping to achieve more accurate product discrimination and to automate the inspection process.

Tofu (bean curd), which is made from curdled soy milk, is a food item that is now drawing attention. Tofu is a typical gelatinous food made by thermally metamorphosing soy protein and insolubilizing it with metal ions such as calcium. The curdling process is simple and there are many types of tofu, all much alike. Today's manufacturing methods and the current product are the result of much accumulated experience and technique. The quality of tofu is generally determined by its firmness and elastic texture, but its physical quality is determined by the curdling process. To evaluate the curdling process, the sequential change in viscosity was measured by use of the torque measurement method. A parameter design was performed to optimize the curdling of tofu with curdling conditions based on quality engineering.