Axle load distribution for mechanistic-empirical pavement design

Mechanistic-empirical (ME) pavement design often demands default or assumed axle load spectrum data. Using single and tandem axles of the Vehicle Class 9 as examples, this study analyzed the spatial and temporal variations of the load distributions from the long-term pavement performance program traffic database. The study found that both spatial and temporal variations are significant; therefore, it may be questionable to use a single default load distribution factor (LDF) for each axle type of vehicle class in design. However, when conducting cluster analysis, the study found that the large amount of data can be classified into a limited number of clusters, from which multiple load distributions can be developed. These distributions, together with spatial and temporal information, may assist engineers in identifying more accurate load distributions for ME design. Based on trial design results using the mechanistic-empirical pavement design guide software, it was found that different LDFs resulted in months of the difference in predicted pavement lives at the same threshold levels for various types of distresses.