Search

The aim of the paper is to present the computationally timeconsuming task of simulating the process of randomly oscillating thin-shell structures to realize an optimal design with limitations in terms of fatigue damage. The most important aim here is to design an effective optimization algorithm and choose an appropriate approach for the evaluation of multiaxial fatigue damage in the context of the random and non-proportional character of stress in the structure when considering the standard finite element model. The authors propose their own optimization algorithm, which is described in the present study and, on the basis of tests, has proven to be suitable for the aforementioned technical problems. The proposition of algorithms for calculating the accumulation of fatigue damage in non-proportional and multiaxial stresses (especially in terms of the application of rainflow analysis by decomposition of the equivalent stress, as determined by the appropriate “fatigue” criterion) is very important in such computational processes. The entire computational process was implemented in MATLAB with the “Discret_Opt_Fat” main control program. The article presents the theoretical basis for the presented complex problem solution, its algorithmization and the technical application.

The article presents a procedure for vibration analysis of the device based on measured data in simulated operating modes. Subsequently, the criterion of fatigue damage is formulated because this mechanism can be induced or accelerated by recognized operating deflection shapes. The criterion is used to select an optimal set of control device parameter values, which control each device’s operating modes and transitions between these modes and affect the excitation of the frame vibration due to possible shocks in the system. The criterion is formulated based on the vibration displacement processing from these dynamic measurements, the modified Goodman and Palmgren-Miner rules, and the results of static computational analyse. This criterion compares the effect of damage to the modes that control these sets of parameters, while the selection of the optimal control set is performed. At the end of the article, the limitations resulting from the simplifications used are described. The application of this procedure is also useful, for example, when further data are obtained by subsequent experimental stress-strain analysis methods and when we want to reduce the number of possible operating modes by which we carry out subsequent measurements among the most damaging modes.