Reducing manufacturing defect through statistical investigation in an integrated aluminium industry

Modern philosophy of quality management enhances thrust on customer satisfaction. Achieving continuous improvement in quality is a way to reach the ultimate goal of customer satisfaction. Statistical process control technique is a well-known analytical technique, which is used to solve quality problems in industry. In this paper we present how this technique was used to solve a quality problem through planned data collection and the use of statistical tool. This study was conducted in an integrated aluminium industry in India who was facing poor customer acceptance of one of their high valued product webstock, which was used to produce toothpaste tubes. Pareto analysis showed that a dragging problem, which resulted in a short length of the toothpaste tube, was the most frequent problem. High and inconsistent coefficient of friction (cof) was identified as the root cause of this dragging problem through planned data collection. A detailed and in depth study was initiated to achieve low and consistent cof. Optimum conditions of the process parameters were obtained using design of experiments viz Taguchi's orthogonal array. The recommendations were validated by confirmatory trials. The desired range of output cof was achieved. The recommendations were implemented as a standard operating practice. As a result of implementation the occurrences of the dragging problem was substantially reduced.