8. Quality and stability analysis of repaired products using check cards

Annotation: The quality analysis of products is an essential component of a production process and a critical factor for reliability enhancement. The continuously increasing complexity of manufacturing processes leads to the need for more eff ective quality inspection methods that would ensure optimal time and resource costs. The statistical process control method is among the most popular approaches. It supports reasonable decisionmaking about product quality based on the results of random checks. The objective of this research is to justify the application of the check card method. A check card is a tool for the qualitative statistical quality control of a production process (in this paper, the object of such control is an incoming inspection of repaired products). Check cards are used to evaluate qualitative parameters for concluding about the stability and quality of the inspected batch. Further, the check card approach enables repair process correction as necessary. The scientifi c novelty of this approach, as the authors believe, is the practicality of correcting inspection results during the process, diff erent from the continuous control method, for instance, thereby managing the process’s quality and stability. The paper investigates the application of a popular statistical control method, namely, the check card method based on quantitative properties. An approach based on checking the quantitative value of a process parameter is applied (in this paper, a parameter that is examined during repair). This approach diff ers from the check card method that uses qualitative properties to verify the average frequency of batch rejections (known as the r-cards). Mathematically, this approach is based on checking the statistical hypothesis that if a parameter demonstrates a normal distribution, a check card is generated within this distribution with a probability approaching a unity, approximately 0.997. The paper presents an algorithm for checking normaldistribution input data using the ω2 criterion. Using actual data acquired during repairs, the paper provides an example of computations and check card generation, along with a results analysis and conclusions. Quality and reliability control professionals, as well as experts of manufacturing enterprises, can use the fi ndings of this research to introduce a system for performing statistical quality control during product repair and manufacturing.