Cement, Energy and Environment

t ----- variance and standard deviation display the variation in the process parameters; the mean, mode and median measure the central tendency of the scattered data. The distribution average or mean is an independent measure, like the standard deviation. Two sets of distributions may have the same average but different standard deviations. Also, they may have the same standard deviation but different distribution averages. It is likely that the center of distribution as measured by the mean may have shifted resulting in some 1tems falling out of the specification limits. When the market is customer-driven, all products must satisfy the specifications as laid down by the customer. Manufacturing process has to be controlled within given limits of variability. Products/items falling outside the specified limits are rejected. Hence the manufacturer takes every care to prevent the risk of rejection. There are two distinct controls applicable to the product; one is the manufacturing process control. which is achievable through the available precision and accuracy of the plant machinery by the Door PM SL PM - Proc~ss mean J>R - Process range SL - Specification limit manufacturer and the other is the customer specification, which concerns the market acceptance of the product. So, there are producer and buyer in this game. The manufacturer must produce what the buyer wants. Hence the manufacturing process variations should be reduced or controlled to a level such that it permits a displacement of the distribution average from the mean to the extent of 1.5 times the standard deviation and still staying within the specified limits of acceptable quality. It means narrowing the process dispersion. It provides a margin of safety or a cushion between the span of process control and the spread of customer specifications such that the process control variations, as expressed in sigma units, must remain comfortably within the customer specified limits. This is the concept of six-sigma approach. Six-sigma is expressed in defects per unit, which can be a load, a design, time frame, distance, money etc. depending on the situation. This concept is illustrated as follows. Door . Design: The door-width in the house comfortably allows the inmates to pass through with sufficient empty space on either side for one person. For example, a 3 feet wide door for the average human width of 1.5 feet, having a standard deviation of 0.25 will allow almost all people to pass through it. There is adequate design margin for safe passage. Here, door width is the specification limit and human width as the process range. Centerline of the door coincides with the process mean. When one rs entering the door, a bit away from the door center line, still there 1s a clear margin for free passage. Process variations can be interpreted by the entry of a thin or a fat person through the door. Exhibit 1 explains the situation. Product characteristics: In cement manufacturing , the fineness of Ordinary Portland Cement is specified under ISO standard, as 2900 to 3050 Blaine (measure of fineness) . Accordingly by setting the separator, the cement mill produces an average fineness of 2975 Blaine with a standard deviation of 12.5 in order to Exhibit- 1 =-iPS'+- SL PS - Process shi 1\ 15