In an initial report, earthquake prediction methods were studied in the Tsukuba area and the possibility of earthquake prediction using the MT system was shown. Based on these results, further studies were carried out in three areas in Ibaraki prefecture(Tsukuba, Hitachinaka, and Juo) to test the possibility of earthquake prediction one hour in advance in each area. Specifically,S/N ratios expressed as items R0 to R15 were obtained by periodic analysis of vibration velocity(μm/s) for six minutes calculated in 1/100-second steps, scaled by 2n. Using these S/N ratio patterns, distances were calculated by the RT method in the MT-system from one hour before to four hours after the earthquake. Changes in these distances were then analyzed by the two-tailed T method, which made prediction one hour before the earthquake possible. The prediction error in each area, calculated using the data obtained in the three areas, was about 43%, or 2 to 3 levels on the Japanese seismic intensity scale. It was demonstrated that the prediction error differs depending on the data processing method and that all the items resulting from occurrence of each earthquake were separate.
A new index that applies the error root mean square method, which can discriminate and diagnose comprehensive pattern changes in multidimensional information, is proposed for assessing the environmental impact of a factory. The following possibilities have been confirmed : (1)In comparison with unit consumption evaluation, this index is less affected by increases and decreases in production, so it is more stable, can lead to quick and accurate environmental improvement actions, and can make such activities more efficient overall : (2)By diagnosing individual items, this method can target actions at a more limited set of specific items than before and motivate improvements in the factory itself, which leads to speedy and voluntary actions. The MT system approach can be applied not only to environmental impact assessments but also to daily activities of companies and other organizations to achieve more efficient operation.
There is concern about the so-called sick house syndrome, a set of adverse health effects thought to be caused by chemicals emitted from building materials etc. Currently, Japan's Ministry of Health, Labour and Welfare regulates in-building concentration levels of 13 chemical substances that cause indoor air pollution. One of these substances is formaldehyde. Coating materials in general undergo some type of hardening reaction, and depending on the type of hardening reaction, there may be a side reaction that produces trace amounts of formaldehyde. A study was made of the reduction of formaldehyde emission from a reactive coating material used for building floors. The precise measurement of trace emissions of formaldehyde has been a major issue in the development of coating materials and the determination of their fbrmulations. As reported in this paper, a model experiment devised for measuring formaldehyde emissions led to an improvement in the accuracy of this measurement, this measurement method was used in an L18 orthogonal array study of the formulation of the coating material, and a formulation that complied with the formaldehyde emission guideline was found.
Usually a mixture of several kinds of coagulants is used in the manufacture of tofu. The subject of this study was a typical mixed coagulant, a blend of magnesium chloride for old-fashioned taste and GDL(glucono-δ-lactone), which simplifies control of the coagulation speed. There was concern that this coagulant might tend to cake, but there had been no major caking trouble for more than ten years up until 2011, when there were a large number of complaints about caking.
Severe caking was recognized in an experiment using an actual sample. In this study, utilizing the reproducibility of the problem of obvious caking, a redesign of the coagulant was undertaken. An attempt was made to solve the caking problem of the mixed coagulant by using the quality engineering methodology of parameter design to clarify the control factors strongly related to caking. Parameter design was performed for the coagulant using estimated caking conditions and tofu manufacturing conditions. Conditions under which a Iong caking time could be obtained were determined from the smaller-is-better S/N ratio. A coagulant that resists caking even at about 40 degrees Celsius, which is thought to be the highest temperature used in Japan, was obtained.