A random stress spectrum online analysis method and its application in fatigue damage analysis of metallurgical cranes

Metallurgical cranes operate under complex and continuous lifting conditions, leading to large volumes of structural stress monitoring data over extended periods. Conventional stress spectrum analysis methods often result in the accumulation of residual waves, reducing accuracy. To address this issue, a novel online analysis method for fatigue damage assessment of metallurgical cranes is proposed. Initially, a step-type rainflow counting method is employed to segment and count real-time generated random stress data, providing real-time output of random stress spectra. When the cumulative residual waves reach a certain threshold during cyclic segmentation counting, a secondary fusion counting of the residual waves is performed. The results are then incorporated into the random stress spectra to correct the real-time counting, thus improving counting accuracy of subsequent fatigue damage assessments. Using S-N curves combined with Miner's damage theory, the step-type rainflow counting is applied to analyze the real-time generated stress spectra and calculate fatigue life, forming an online analysis method tailored for fatigue damage assessment. Simulation experiments are conducted using both four-point rainflow counting and step-type rainflow counting methods on random stress spectrum data. The results show that the accuracy of step-type rainflow counting reaches 99% of conventional rainflow counting. This online analysis method can be applied to analyze stress data from a measurement point at the connection of the sub-main beam of a metallurgical crane. The effective fatigue life at this measurement point is calculated. The results demonstrate that this online analysis method effectively enhances the real-time assessment of metallurgical crane fatigue, providing reasonable evaluation results and a timely reliable basis for long-term crane maintenance.