Graphical rainflow counting : complete slgorithm for single courses (SCRF)

Abstract

Rainflow counting can be considered the most important method in counting random stress- or strain histories with respect to time in the field of fatigue strength. The method is often explained by rules of a rain flow which gave the name to the method. The algorithmic implementation nevertheless is usually done in three-point or four-point algorithms which work by comparing the differences between the adjacent reversals and then delete the points of the identified intermediate cycle from the time history.

The SCRF-algorithm presented in this contribution determines the courses of the rain flow in terms of their start und stop classes as half-cycles. In a subsequent step the half-cycles are matched together to form a closed loop resp. kept as a residue if the counter-course is missing. The algorithm is explained in general und by means of an example. For verification the SCRF-algorithm is compared to a three-point and a four-point algorithm.

For periodic signals, there is complete agreement between the three methods. For non-periodic signals, there are small differences: The SCRF can combine some more half-cycles into closed cycles which remain as residues in the other methods. Furthermore the sense of direction of the hysteresis loops is captured more precisely with the SCRF due to its chronological approach.