Statistiese beheer van die gewigte van sementsakkies in 'n verpakkingsaanleg

Abstract

STATISTICAL CONTROL OF THE WEIGHTS OF CEMENT BAGS IN A PACKING PLANT. 1. Introduction - Alpha Cement Dudfield is a cement producing facility situated ± 25 km west of the North West Province town of Lichtenburg. It has a production capacity of over the 5000 tons of clinker per day and produces an average of 35 000 bags of 50 kg cement per day. Due to this large amount of bags produced on a daily basis slight weight deviations in the average weight of the bags may result ,over a long period, in a significant loss or gain of cement. This loss or gain of cement may mean a profit or a loss to the company and the client. To ensure continuous consumer satisfaction the loss or gain of cement can be minimised by proper control of the weights of the cement bags produced. A Rotary packer equipped with accurate electronic weighing equipment precisely measures the bag weight of each bag that it produces. By applying probability theory we aim to ensure the optimum production rate of this packer within given standards. 2. Objective - The objective of the project is the application of the normal distribution to a cement bag filling process in order to control the weights of the bags within certain parameters. This includes a literature study on probability with the focus on the normal distribution, a literature study on decision making using the normal distribution, a literature study on process control using the normal distribution and a case study on the control of an eight spout cement packer using the theory discussed. 3. Summary of results - The SABS tolerances specified per bag of cement are between 150 grams underweight and 250 grams overweight. To ensure that these tolerances are met, continuous control of the bag weights are necessary. By analysing the bag weight data by means of the normal distribution, useful information regarding tonnage gained or lost can be estimated. This information can also be used in a reactive way by setting up control charts and by reacting early on warning signs given. The standard deviation and average for 1996 and 1997 are determined and discussed. Possible mechanical deviations on the packer are identified as well as ways of overcoming them. 4. Conclusion - By correctly using the normal distribution to identify possible problem areas in the control of the weights of cement bags, the amount of cement lost by the company and the customer can be minimised. This not only leads to satisfied customers but to long term financial gain for the company.