PSI - Issue 13

Petr Lehner et al. / Procedia Structural Integrity 13 (2018) 1539–1544 Lehner P., K ř ivý V., Krejsa M., Pa ř enica P., / Structural Integrity Procedia 00 (2018) 000–000

1540

2

1. Introduction The phenomenon of fatigue is one of the major factors influencing the service life of steel structures supporting overhead cranes. Correct prediction of the service life of supporting structures is essential for reliable design of structures exposed to fatigue loading. An important engineering problem is also an estimate of the residual fatigue life of the existing constructions. For these reasons, it is very important to develop methods for calculating and evaluating residual fatigue life and time-dependent reliability analysis of existing steel structures (Xiang and Liu, 2011; Ye, Su and Han, 2014). Existing constructions were implemented according to the rules and procedures in force at the time of design. The design of constructions made in the first half of the 20th Century usually did not include the assessment and prediction of fatigue life (Zhang and Mahadevan, 2001). Currently, engineers can use advanced computational techniques based on numerical modeling and a stochastic approach to assess fatigue resistance. Several studies of the stochastic estimation of life over time of structural details or systems are available (Kala, 2015; Vičan et al. , 2016; Ghosh et al. , 2017; Králik, 2017). However, there is currently no comprehensive probabilistic methodology generally applicable to structures supporting overhead cranes. Similarly, there is research on cracks in the steel. Propagation of fatigue cracks and possible predictions of such propagation over time from the beginning of variable load effects is the case where the probability method has to be used because too much uncertainty affects the determination of input values (Krejsa et al. , 2017; Seitl et al. , 2017). However, in the analysis it is necessary to determine the location at the structure in which such a crack could arise. It was required to create a complex static 3D model with a moving load. Results from load analysis were used for indication of the worst element of the structure. However, realistic input data about stress load history in selected structural elements was necessary to determine first. The direct calculation of the time-dependent load effects is a possible approach to solve this problem. In many cases it is the only option to obtain the necessary input data for fatigue assessment. The accuracy of the chosen calculation procedure depends on many input factors, which are obviously variable (e.g. weight of lifted weight loads, position of the overhead crane and crab, number of crane crossing per operating cycle). In the next step, the sorting of load parameters using the Rain Flow method was used. Sorted data from the Rain Flow method can be applied to the FEM model of the structure to evaluate the resulting fatigue resistance of structural elements. The calculation methodology is explained on the existing steel structure that was built in 1920, and which is still in operation (see Fig. 1).

Fig. 1. Overall view of the examined construction.

The selected construction is a typical representative of the historical industrial structures exposed to cyclic loading, for which their operator requires verification of their residual life. Also there were possibilities of real material testing and extensive documentation including many inspections and expert information.