Lifetime tests and prediction of the usable remaining service life of rubber-metal elements

Condition monitoring is already used in various technical applications due to its many advantages. It was investigated to what extent condition monitoring based on the developed methods is able to estimate the remaining useful life of rubber-metal elements. These elements are used for vibration engineering tasks, especially for reducing continuous vibrations, reducing peak loads during resonance passages, shifting of resonance frequencies and vibration decoupling. A test bed has been set up for accelerated lifetime tests of rubber-metal elements. Various sensors are integrated in the test bed to record various lifetime data of rubber-metal elements. This data is evaluated within prediction methods.

The relative temperature has proved to be a suitable variable for monitoring the service life of rubber-metal elements. The temperature of the element is set in relation to the ambient temperature and recorded during the tests. Since in real applications non-stationary operating conditions occur, i.e. operating conditions that vary over time, life tests were carried out under stationary and non-stationary operating conditions. Frequency and amplitude of the force as well as ambient temperature were varied over time. Based on the measured data, the remaining useful life of the elements was estimated by a suitable prediction method, a multi-model particle filter. Based on multiple models various uncertainties, such as operating conditions and the uncertainty of the degradation process, were taken into account. Besides a reliable and accurate prediction method, a failure threshold was defined. This threshold characterizes the end of life of the elements or the state in which the elements are to be replaced in real applications. The failure thresholds are calculated methodically in the absence of expert knowledge. The project’s aim as to achieve an efficient and precise estimation of the remaining useful life to generate the basis for a predictive maintenance planning.

Publications of the chair on this research area can be found here.