Clustering Model for Correlated Load Processes

Many loading on structures (such as wind, wave, snow, and rain-on-snow) may be due to a common source and have a tendency to cluster around common points in time so that the occurrence dependence may be an important factor in load combination analysis. Based on a multivariate point process, an analytical model is presented to take such occurrence dependence into consideration. Methods for analysis of mean rate of load coincidence and probability of combined load are developed as an extension of the load-coincidence method previously proposed. Approximate solutions are obtained in simple, closed form. The accuracy of the method is verified by Monte-Carlo simulations. The load coincidence rate is extremely sensitive to occurrence clustering. As a result, when such dependence exists and is not considered, the estimate of the probabilty of combined maximum, as well as the reliability of structure could be seriously in error.