Abstract
The method of finding the optimum conditions of operation of machinery, outlined in this paper, was originally developed for use in the chemical industry particularly on production, where only small changes in the variables can be made at any one time, but where the response can be measured very precisely. In cotton spinning, the measured response—the coefficient of variation of thickness or yarn strength—is a much more variable quantity and in many cases approximate optimum conditions are known. In such cases we have found it useful to apply the procedure as developed for the study of near-optimum regions to the region about an estimated optimum point and to obtain an approximate contour plan of the response over most of the practical region. In other cases where the experimental field is large, we have used the method of steepest ascent to approach the optimum region.
The procedure in both applications is greatly simplified by the use of standardized variables and suitable experimental designs. The subject of experimental design is covered elsewhere and is only briefly touched upon in this paper.
The examples given in the text show that the results are readily interpreted and enable several types of question to be answered. The conclusions that can be drawn from such a set of results are of immediate and practical use.