Crossing-consistent peak count The most conservative crossing-consistent cycle count

Abstract

This study introduces a new variant of the peak count, termed the crossing-consistent peak count. Crossing consistency implies that the number of cycles with peaks above and valleys below a certain level equals the number of up-crossings of that level. The conventional peak count does not exhibit crossing consistency. In the crossing-consistent peak count, valleys are paired with the nearest peaks, from the largest to the smallest, and each cycle is counted as a full cycle. To include half cycles, the reversals of the time series are considered twice, except for the start and end points, such that each cycle is counted as a half cycle. Based on the fatigue framework developed by the Lund University, in which the cumulative cycle histogram is a key element, this study shows that the crossing-consistent peak count is the most conservative crossing-consistent counting method. This holds true for all time series, not only Gaussian signals. The crossing-consistent peak count was applied to wind turbine loads and compared with the rainflow count and conventional peak count. Moreover, it was applied to broad-band Gaussian signals, and it was concluded that the mean damage follows the narrow-band approximation.