Virtual design has often been employed in past usability evaluations; results obtained by applying virtual design to a mechanical design are reported here. Virtual design was applied to the conceptual design of a dust shutter provided in a multifunction electrophotographic printer to improve the immunity of its image density sensor to fouling. Four evaluators, of whom three were well acquainted with the subject and one was not, made a sensory evaluation of the fouling of the sensor surface on a seven-grade scale in which fouling was treated as a smaller-is-better characteristic, and the proposed conceptual design was evaluated with respect to two types of fouling having different causes, derived from field reports of existing printers. The factor effects obtained as a result differed from one evaluator to another, but reproducibility was obtained for substantially all of them. On the basis of this all four evaluators conducted a design review, and were able to select the best proposed design on the basis of concrete, quantitative information, specifically, a factor effect diagram. An improvement effect was also obtained in a confirmation test carried out on an actual printer. During the process, information and findings that appear useful for the future development of virtual design were obtained, such as the zoned grading method, selection of a proposed design through consultation, and the treatment of knowledge.

A quality engineering study was carried out to increase the precision of V-bending with a hydraulic press brake, using commercially available press brake tooling. Experiments were performed with a novel test piece having three types of holes positioned on the bending line to interfere with the precision of the bend. After bending, the straightness of three lines on one side of the bend was measured,treating deviation from the ideal position determined from the endpoints of each line as a nominal-is-best quantity with zero as the nominal value. Unbalance between the pressure applied on the right and left sides of the press brake was taken as a noise factor. In the first experiment, eight control factors were selected from among the machining conditions, blank conditions, and tooling conditions and the selected conditions were assigned to an L₁₈ orthogonal array. After optimum conditions were determined from this experiment, a second experiment was carried out and a high level of reproducibility was confirmed. When the optimum conditions were applied to actual parts under development,high precision was obtained from one-shot forming. A further advantage was that since commercially available tooling was used, tool procurement time was nearly nil and the tooling cost was reduced by a factor of ten, so development time was greatly shortened and development cost was reduced considerably.