Usefulness of a statistical analysis in the quality control of plastics for household appliances

In this work, the problem of quality control of plastics destined to household appliances is faced. The material is poly-propylene charged with mineral filler, drawn from two batches, each provided by a different producer. From each batch, dog-bone shaped specimens for tension testing were fabricated by injection moulding, using two different injection rates. The samples were tested in tension up to failure, and the tensile strength and Young’s modulus were measured. In all, sixteen tests were carried out for each experimental condition. From the mean values of the Young’s modulus and strength, the difference between the experimental conditions adopted could not be simply appreciated. However, the application of the analysis of variance (ANOVA) clearly revealed the effect of both the material producer and the process conditions, especially when the strength was adopted as a parameter. In order to model the strength and modulus distribution, a two-parameter Weibull curve was assumed for the mechanical properties measured. From the results, the Young’s modulus and tensile strength reasonably follow a Weibull curve. Moreover, also the parameters of the curve proved to be capable to discriminate between both the different batches and the different process conditions.