抄録
Robust design is crucial to maintain the function of the design object because it ensures function stability (hereinafter called robustness), and many methods (hereinafter called RDM) were proposed. These methods enable the products to ensure robust performance in diverse users and the surroundings has received much attention. Additionally, the design with ease for user of use and reducing the cost can be implemented. However, this method fixes the value of the control factor, the range of the Probability density objective characteristic is large compared to the tolerance, resulting in low robustness. Herein, in this research, a robust design method increasing the robustness of the object is suggested. The suggestion of this new method is realized by expanding controllable factors, which were set to a fixed value, to variable control factors, which can be changed by users through usage, resulting as keeping the target quality distribution within a permissible range. In order to prevent the increase in parts and cost caused by employing variable control factors, and to prevent accidents caused by miss usage, the proposed method derives the minimum range of the variable region through the following steps. First, analysis to evaluate if robustness is ensured even if the variable control factors are set to a fixed value, by setting the standard deviation, calculated by the combination of each variable control factors, as the evaluation measurement. Next, Analysis to evaluate the robustness even if multiple variable control factors are connected, through setting the share of the oval and brachytype's area that implicates all evaluation points. Finally, based on these analysis, the most suitable range of the variable region is drerrived.
