No abstract

No abstract

In this work, a methodology that uses the dynamic Bayesian networks (DBNs) in combination with an idea algebra is developed for assessing the dynamic reliability of engineering systems. A network representation of the system topology is first introduced in the form of “idea” objects representing components and their functional interfaces, thus integrating the functional and material descriptions of the system. Various time-dependent functionalities can thus be mapped to segments or loops of the resulting network, which are then translated automatically into the form of a DBN, thereby avoiding the need to manually generate the dynamic fault tree (DFT) logic that would normally serve as a starting point. The methodology is demonstrated in a case study, where reliability analysis of an automobile system is performed. The idea algebra is automatically deployed in Mathematica and evaluated in the GeNIe platform. Weibull distribution was used for the generation of the dynamic values for the reliability analysis of the system within a certain period.

In this work the quasi-static model of the three-dimensional geometrical non-conjugate contact problem for two [Formula: see text] surfaces is studied. The set of contact equations is formulated by using a new parameterisation that enables to reduce the conventional system of five nonlinear equations with five unknown position and contact parameters to just two nonlinear equations with two changeable parameters. The novel model is computationally efficient and demonstrates increased accuracy and stability of the numerical solution, compared to the conventional model described by Litvin, which suffers from convergence problems and requires a high computational effort. The new model is implemented to spur gear with crowned tooth surfaces to parametrically estimate the susceptibility to diverse misalignments of the contact pressure, transmission error and path of contact.

No abstract