Portal de Periódicos da CAPES

Quality selection of a two-dimensional process level under single acceptance sampling by variables

1992; Elsevier BV; Volume: 27; Issue: 1 Linguagem: Inglês

10.1016/0925-5273(92)90125-q

1873-7579

Olle Carlsson,

Industrial Vision Systems and Defect Detection

The present study has its origin in quality problems faced in a hardboard mill. The quality control of hardboard includes the measurement of several quality characteristics, where special attention is paid to the thickness and the density. Samples of sheets are taken from hardboard lots and a lot is accepted if both the mean thickness and mean density exceed given levels. If not, the lot is rejected and sold at a reduced price. The producer can reduce the risk of rejection by making a thicker and heavier board. This will increase the cost of production and thus render a decrease in net income. The paper presents a model to determine the optimal two-dimensional process level and the optimal expected net income per lot. The model is based on assumptions of a known process variance and of a net income per lot, which is a linear function of the quality characteristics for accepted and rejected lots, respectively. An approximate optimal two-dimensional process level is discussed and an example is given.