Markov matrixes for random sequences imitation in fatigue testing and design in the problem of safety of steel structures

Abstract

To guarantee the safety of steel structures it is important to deal properly with the problem of representation of exploitation loading. For metal fatigue testing and design, the proper choice of random (irregular) loading type is very important. The principles of random loading are discussed and some alternative approaches with their pros and cons are shortly reviewed in the paper. As a sound decision for random, but taking into account some specific features of the exploitation loading process, the target Markov method is proposed. According to this method, the important information of the real random processes in the form of the turning point is used for filling the square Markov matrix (analysis phase) and later on, with employing the random number generator, serves as a source for creating of the so-called replicas. The replicas are the random trial for numerical estimation of longevity scatter. Due to the fact, that all these manipulations are performed with the aim of metal fatigue investigation, some important processes’ characteristics for fatigue, like irregularity factor, fullness factor and machine part longevities were compared. Some important suggestion for the future development of this method, that is taking into consideration the sequence of the events, is discussed.