Physical model based reliability analysis
In a fuel supply system of a conventional household oil heating system the fuel degrades over time depending on the system temperature, the tank volume and the flow rates. The test fuel is a blend of 80 vol.% conventional domestic heating oil and 20 vol.% fatty acid methyl ester from rapeseed oil feedstock. The fuel will increase the deposits in the fuel supply which lead to system failures. In this paper a model based approach is introduced for the analysis of accelerated life testings. The impact of the tank volume on the failure time is analyzed. A well balanced physical model is set up for stochastic simulation. The model contains short term dynamic and long term transient degradation effects that have an important impact on the system reliability. The degradation effects of the components and the used fuel cause a system failure over time. A system failure is detected when the volume flow of the fuel supply declines. The analysis shows that the typical variances of characteristic parameters lead to a significant distribution of failure time. Furthermore, the functional relation between tank volume as stress factor and failure time is described.