Statistical modeling of the squeak noise occurrence of natural rubber

In this study, the occurrence of the squeak noise according to various material / environmental factors of natural rubber was found through friction test using custom built friction tester. Material factors such as rubber hardness, surface roughness and additive content, and environmental factors such as heat aging, temperature, surface moisture, friction speed and frictional load are considered. And the hypothesis test was conducted to determine whether there was a significant correlation between each factor and the occurrence of squeak noise. Afterwards, logistic regression analysis and neural network analysis, which considers the interactions among these factors, were performed using the statistically significant factors in the occurrence of squeak noise obtained by the hypothesis test, and a probability model of the occurrence of the squeak noise was developed. And we compared the occurrence of the squeak noise predicted by the two models with the actual occurrence of the squeak noise, and the result shows 88.15% and 87.16% of accuracy, respectively. We also verify the accuracy of two models by using receiver operating characteristic curve (ROC curve), which illustrates the diagnostic ability of a model.