Resumen

In the production plant of Diana Corporación S.A.S. in the municipality of El Espinal in the department of Tolima, the Diana brand rice is produced. For the company it is important that its finished product complies with the established quality standards. The percentage of broken grain that the units of finished product possess is the most important quality characteristic because it is this one that gives the commercial value to the product. For this reason a capacity analysis study was carried out to determine the specification limits that must be had for Diana brand rice, taking into account the quantity of accepted units, since if quality policies are not complied with, spaces can be generated for potential sanctions and the legal requirements and customer satisfaction are notoriously affected. The development of the work was carried out in 2 hoppers of the packing area with different design; these hoppers handle a flow of channeled type which is not recommended for the process they handle. The study was performed 3 times on each hopper, where approximately 32 samples of finished product were taken and 8 during the filling of each hopper. These samples were taken to the laboratory of the production plant to determine the percentage of broken grain that had one, these data were recorded in a format that was created during the collection of samples. It began by using Shewhart's control charts to verify that the data collected was under statistical control, the charts were used with control limits of ±3σ. Those data which exceeded the control limits were removed from the study and an analysis was performed describing the actions taken during the collection and analysis of that sample. After verifying that the data were under statistical control, a series of hypothesis tests were carried out to determine if the replicas made of each hopper were statistically similar in mean and in mean. In the case of a deviation, it was found that one of the replicas did not coincide statistically with the other replicas, so this one was considered a separate hopper. Those hoppers that were statistically similar in mean and deviation were joined together to create a single control chart, perform a general capacity analysis and be able to determine the acceptance limits for Diana brand rice. These were taken into account at the time of carrying out the capacity analysis, because one requirement for this analysis is that the data behave normally. In such a case, if they were not normal, a goodness of fit test had to be carried out to identify the type of distribution of the data, the data of the replicas had a normal distribution but at the moment of joining the replicas that were statistically similar and performing the normality test again it was found that the data did not follow a normal distribution, these data were transformed by means of the Box-Cox transformation. Finally, a capacity analysis was performed on each hopper studied and on the union of the replicas that statistically coincided in mean and deviation, it was verified that none of these gave acceptable results in the capacity analysis. It was necessary to identify the intervals of tolerance, these intervals had to comply with a minimum of accepted units, after having the intervals of tolerance the analysis of capacity was carried out again only for the union of the replicas to determine if these intervals were acceptable for the finished product. It was concluded that the limits of acceptance of percentage of split grain that were determined increased the number of units accepted in its process but decreased the demand on the quality of the product, that is why it was recommended to Diana Corporación S.A.S to change the design of the packaging hoppers for a design with mass flow and to carry out the study again, since with the current design it is not possible to meet the quality requirements.